diff --git a/docs/reference-matrix.md b/docs/reference-matrix.md
index 4be14b07c..9610d8106 100644
--- a/docs/reference-matrix.md
+++ b/docs/reference-matrix.md
@@ -444,6 +444,7 @@ the candidate mechanisms.
| --- | --- | --- |
| [Issue #60 report](https://github.com/InferMatrixLab/InferRSI/issues/60#issuecomment-6310920198), 2026-10-21; eight Git source inventories in the [owner metadata](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json) | Task/quality/workload separation across text, agentic coding, work agents, TTS, full-duplex interaction and image/video; selected pinned AIPerf/realtime/FDB/video sources rechecked and the current InferRSI metric contract inspected | Extend [inference-benchmark-design](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md) with [task selection](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/task-benchmark-selection.md) and [command templates](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/benchmark-commands.md). Candidate guidance; source review and bounded smoke do not qualify all tasks or establish a performance gain. Replay scores, reference parity and task quality retain separate meanings. |
| [NVIDIA/NeMo-Speech.cpp @ 20999cc9](https://github.com/NVIDIA/NeMo-Speech.cpp/tree/20999cc921d2b14aa58708f61756a8b2e787ba20), inspected 2026-11-01 | Selected cache-aware encoder/config, runner, SDK bounds, endpoint policy, geometry test/CMake and ASR benchmark; complete 14-file readset and Apache-2.0 / parakeet MIT attribution retained in the reference | Extend [inference-benchmark-design](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md) with [cache-aware ASR stream boundaries](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/cache-aware-asr-stream-boundaries.md): first/steady mel geometry and decoded clock, stream-slot reset/reuse, EOU/EOF/reset and model-test SKIP boundaries. Source-derived candidate / empirical not_run; benchmark figures remain author-reported with precision, FE and trial limits. No model/device/quality/long-audio/load/service qualification. |
+| [Embodied.cpp repo](https://github.com/SEU-PAISys/Embodied.cpp/tree/156bed151826780e2ce0132f8d0e3540889ae8cd) + [paper v3](https://arxiv.org/abs/2607.02501v3), inspected 2026-11-04 (repository Apache-2.0; paper CC BY 4.0) | Selected fixed CPU-observation-to-complete-CPU-action boundary, client queue/profiler denominators, precision identity, per-task quality and source/artifact lineage; [exact readset and rights](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-source.json). Later SmolVLA reports remain separate from the August paper. | Extend registered owner `inference-benchmark-design` ([SKILL.md](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md)) with [VLA boundary and quality guidance](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-boundary-and-quality.md). Source-derived candidate/not_run; public sample arithmetic is distinct from episode recount and native execution. Generated actions are not control frequency; per-task loss, unequal precision, whole-device memory and unresolved measured-script/dirty-patch identity remain explicit. |
## SGLang versioned entry selection
diff --git a/docs/searchspace-matrix.md b/docs/searchspace-matrix.md
index aad404af3..d7bfefb37 100644
--- a/docs/searchspace-matrix.md
+++ b/docs/searchspace-matrix.md
@@ -614,3 +614,16 @@ NOASSERTION and InferRSI's repository license.
| Accelerator / RAM / Host | Diagnose / configure / verify: tensor/alias/gradient state, training and data ownership, compiler guards and export contracts, allocator accounting, stream dependencies and graph input/output lifetimes | [pytorch operating graph](../src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/SKILL.md), [memory/graph route](../src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/sub-skills/memory-streams-and-graphs/SKILL.md), [compiler/export route](../src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/sub-skills/compiler-and-export/SKILL.md) | Source version `2.16.0a0` and observed CPU wheel `2.15.0+cpu` are distinct. CPU observations do not qualify pinned-source execution, CUDA/ROCm/MPS/XPU, native custom operators, memory fit or model quality. |
| SSD / Storage / RAM | Diagnose / verify: serialization trust and state identity, checkpoint ownership, distributed restore and capture/staging/upload completion boundaries | `pytorch`: [distributed/checkpoint route](../src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/sub-skills/distributed-and-checkpoint/SKILL.md); existing [DCP async-save owner](../src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/SKILL.md) retains its separate source pin | No source-build checkpoint run, backend durability, distributed recovery or measured storage benefit is established. A state-dict round trip in a separate installed wheel does not qualify those properties. |
| Communication / Network / Accelerator | Diagnose / verify: process/rank/device mapping, backend and collective schedule, DDP/FSDP2/DTensor state and completion | `pytorch`: [distributed/checkpoint route](../src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/sub-skills/distributed-and-checkpoint/SKILL.md), [source and runtime boundary](../src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/references/repo-provenance.md) | Selected source and CPU controls do not establish NCCL or other accelerator transport, multi-device/multi-node correctness, overlap or throughput. Required backend checks remain `BLOCKED_REQUIRED_BACKEND`; no new measured/green resource cell, OS control or compatible latest PyTorch/TorchNPU stack is asserted. |
+
+## Embodied.cpp benchmark methodology
+
+The [inference-benchmark-design](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md)
+method uses [Embodied.cpp at 156bed15](https://github.com/SEU-PAISys/Embodied.cpp/tree/156bed151826780e2ce0132f8d0e3540889ae8cd)
+and [arXiv:2607.02501v3](https://arxiv.org/abs/2607.02501v3) as source context
+for observation-to-action timing, generated versus replayed actions, precision
+identity and per-task closed-loop quality. Its [operating reference](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-boundary-and-quality.md)
+and [readset](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-source.json)
+retain the later author's measurement and source-version limits.
+
+These are comparison methods. Experiment hardware is source context rather
+than a device prerequisite or a resource mechanism for this method.
diff --git a/docs/skill-coverage.md b/docs/skill-coverage.md
index 50c3f1b6c..5a5f7d520 100644
--- a/docs/skill-coverage.md
+++ b/docs/skill-coverage.md
@@ -4,7 +4,7 @@ Generated from the registry, the skill sidecars and the two maintained
matrices. **Do not edit by hand**; regenerate after any registry, skill,
matrix or frozen-plan change.
-- Input digest: `sha256:7640b171256091efd1514a33d5e434eef3feaed238fc92fe3eeba02fee94e9ff`
+- Input digest: `sha256:2ade3ef0deed6fb22ff97c62064e29198e13f8f6bf5e750776cb205d8c14ad43`
- Regenerate: `python tools/build_skill_coverage.py`
- Verify without writing: `python tools/build_skill_coverage.py --check`
@@ -131,7 +131,7 @@ optimization is fully covered or that any procedure has measured benefit.
| [choose-parallel-training-plan](../src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/SKILL.md) | common | Compare stage, device-mesh and operator-sharding candidates for static distributed training while accounting for ... | Comparing stage, device-mesh and operator-sharding candidates for static distributed ... | failure_modes, input_contract, produced_artifact, refusal_point
3 advice | 1 Accelerator · Diagnose / Configure / Optimize / Verify
4 Network · Diagnose / Configure / Optimize / Verify | 1 · Systems sources (Alpa — [2201.12023v3](https://arxiv.org/abs/2201.12023v3)) | candidate · not_run · 1 readset(s) | no-eval |
| [dense-performance-evidence](../src/inferrsi/data/skill_library/methods/dense-performance-evidence/SKILL.md) | common | Taxonomy of performance evidence an optimization agent needs, ordered by collection cost and diagnostic power. Maps ... | Deciding which performance evidence to collect next and which claim each evidence level ... | failure_modes, input_contract, produced_artifact, refusal_point
4 advice | 1 Accelerator · Observe / Profile / Verify
2 RAM/Host · Verify
4 Network · Verify | 1 · Systems sources (NVIDIA/TensorRT-LLM @ f33e0e5fd)
2 · Methodology and corpus references (JiusiServe/vllm-evolve @ 43cc82c)
2 · Methodology and corpus references (microsoft/SkillOpt @ 343db22) | candidate · not_run · 0 readset(s) | — |
| [evaluate-elastic-serving-placement](../src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/SKILL.md) | common | Compare migrating an LLM request between model replicas with changing its sequence-parallel execution group, including ... | Comparing migration of a running LLM request between compatible model replicas with ... | failure_modes, input_contract, produced_artifact, refusal_point
3 advice | 1 Accelerator · Diagnose / Configure / Optimize / Verify
2 RAM/Host · Diagnose / Configure / Optimize / Verify
4 Network · Diagnose / Configure / Optimize / Verify | 1 · Systems sources (Llumnix — [2406.03243v1](https://arxiv.org/abs/2406.03243v1)) | candidate · not_run · 1 readset(s) | no-eval |
-| [inference-benchmark-design](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md) | common | Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents ... | Selecting task quality gates, reference parity checks and performance workloads for text ... | failure_modes, input_contract, produced_artifact, refusal_point
5 advice | N/A: SKILL.md §§ Select the task and its evaluation contract, Statistical requirements, Shared-machine measurement discipline, and Search integrity and final-state verification design comparable evaluation; hardware identity and allocation isolation are comparison controls, not a hardware-resource mechanism. | 2 · Methodology and corpus references (JiusiServe/vllm-evolve @ 43cc82c)
Task-specific benchmark selection (issue #60) ([Issue #60 report](https://github.com/InferMatrixLab/InferRSI/issues/6 ...)
Task-specific benchmark selection (issue #60) ([NVIDIA/NeMo-Speech.cpp @ 20999cc9](https://github.com/NVIDIA/NeMo-Spe ...) | candidate · not_run · 7 readset(s) | — |
+| [inference-benchmark-design](../src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md) | common | Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents ... | Selecting task quality gates, reference parity checks and performance workloads for text ... | failure_modes, input_contract, produced_artifact, refusal_point
6 advice | N/A: SKILL.md §§ Select the task and its evaluation contract, Statistical requirements, Shared-machine measurement discipline, and Search integrity and final-state verification design comparable evaluation; hardware identity and allocation isolation are comparison controls, not a hardware-resource mechanism. | 2 · Methodology and corpus references (JiusiServe/vllm-evolve @ 43cc82c)
Task-specific benchmark selection (issue #60) ([Embodied.cpp repo](https://github.com/SEU-PAISys/Embodied.cpp/tree/15 ...)
Task-specific benchmark selection (issue #60) ([Issue #60 report](https://github.com/InferMatrixLab/InferRSI/issues/6 ...)
Task-specific benchmark selection (issue #60) ([NVIDIA/NeMo-Speech.cpp @ 20999cc9](https://github.com/NVIDIA/NeMo-Spe ...) | candidate · not_run · 9 readset(s) | — |
| [inference-optimization-casebook](../src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md) | common | How to structure and use a casebook of past optimization decisions as a decision-support tool. Extracts the casebook ... | Structuring and querying a casebook of past optimization decisions as decision support ... | failure_modes, input_contract, produced_artifact, refusal_point
4 advice | 1 Accelerator · Optimize | 1 · Systems sources (NVIDIA/TensorRT-LLM @ f33e0e5fd) | candidate · not_run · 0 readset(s) | no-eval |
| [kernel-writing-guide](../src/inferrsi/data/skill_library/methods/kernel-writing-guide/SKILL.md) | common | Decision guide for writing custom inference kernels: when to write one, which framework to choose (Triton, CuTe DSL ... | Deciding whether to write a custom inference kernel, routing a kernel pattern to Triton ... | failure_modes, input_contract, produced_artifact, refusal_point
6 advice | 1 Accelerator · Optimize / Diagnose / Configure / Verify | 1 · Systems sources (Dao-AILab/causal-conv1d @ `cd81f0413cad2fc1e6f17e785ac39f59aae690cd` ...)
1 · Systems sources (NVIDIA/TensorRT-LLM @ f33e0e5fd)
1 · Systems sources (huggingface/transformers @ `0dad7b822255a0ae261ec45ae937371e859ffd1a` ...)
1 · Systems sources (mindspore-ai/akg @ `5aa15f3fe4d856eb60cd3b27447e04f5f1fb4dff` ...)
1 · Systems sources (sgl-project/sglang @ `16176278d0180f1a1a578e53e6eef432577876c2` ...)
1 · Systems sources (vllm-project/vllm @ `537a79f0fc5ba564d2bef43d54915b2f80f9f77d` (root ...) | candidate · not_run · 7 readset(s) | no-eval |
| [optimization-search-policy](../src/inferrsi/data/skill_library/methods/optimization-search-policy/SKILL.md) | common | Policy for how an optimization agent should select, order, and stop optimization experiments. Covers candidate ... | Selecting, ordering and stopping optimization experiments: candidate generation ... | failure_modes, input_contract, produced_artifact, refusal_point
4 advice | 1 Accelerator · Optimize | 2 · Methodology and corpus references (JiusiServe/vllm-evolve @ 43cc82c)
2 · Methodology and corpus references (RRSI — arXiv 2609.24972v3)
2 · Methodology and corpus references (arXiv 2608.01867 (CRISP))
2 · Methodology and corpus references (microsoft/SkillOpt @ 343db22) | candidate · not_run · 9 readset(s) | no-eval |
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md b/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md
index 86ee95d7e..8095c66b3 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/README.md
@@ -15,7 +15,7 @@ These are knowledge abstracts (`--kind knowledge`), with namespaced package
IDs distinct from explicitly selected `checkout_ref` skills. Ascend/CUDA
procedures use their corresponding backend; shared methods use `common`.
-Source snapshot: [7aa3344bf84c42467f43b3a927e1c7d13642bd55](https://github.com/InferMatrixLab/InferRSI/tree/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library).
+Source snapshot: [d5fc2121d6d1628111403ba70ec3b70638ab8674](https://github.com/InferMatrixLab/InferRSI/tree/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library).
```bash
inferrsi knowledge search 'profiling' --repo-type inferrsi --limit 50
@@ -29,7 +29,7 @@ contain the current input paths and bytes (Git checkout filters are respected).
```bash
python tools/build_skill_index.py --source-revision
-python tools/build_skill_index.py --source-revision 7aa3344bf84c42467f43b3a927e1c7d13642bd55 --check
+python tools/build_skill_index.py --source-revision d5fc2121d6d1628111403ba70ec3b70638ab8674 --check
```
`--check` detects missing, stale and unexpected generated records without writing.
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json
index d628bdadd..1dcfc4b1b 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-aclgraph-capture-replay.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-aclgraph-capture-replay/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json
index f4b372d19..7c9ef445e 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-device-health.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-device-health/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json
index 4b940539c..9f4cf0806 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-distributed-parallel.json
@@ -27,8 +27,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-distributed-parallel/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json
index 5c1c8590b..7369e4179 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-graph-mode.json
@@ -26,8 +26,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-graph-mode/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json
index 971e9813a..d82252174 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-model-adaptation.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-model-adaptation/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json
index e8d595cbc..a56c68f2f 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-profiling.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-profiling/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json
index 64816a247..790b28798 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-operator-tuning.json
@@ -24,8 +24,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-operator-tuning/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json
index 7521c9f69..2c3584131 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-optimization-casebook.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-optimization-casebook/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json
index 2a4a0ace2..633e3176e 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-performance-analysis.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-performance-analysis/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json
index a2d677d4d..732bc519c 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascend-video-diffusion-optimization.json
@@ -26,8 +26,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascend-video-diffusion-optimization/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json
index ea618b662..3fa9c0fca 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/ascendc-tiling-design.json
@@ -24,8 +24,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/ascendc-tiling-design/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json
index ca740b5b7..fc8bfcc94 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/cann-attribution-overlap.json
@@ -24,8 +24,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/cann-attribution-overlap/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json
index 7c1cde5ed..6fa679331 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torch-npu.json
@@ -20,8 +20,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/SKILL.md",
"src/inferrsi/data/skill_library/frameworks/torch-npu/torch-npu/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json
index d6676b490..2f1d8b98c 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-device-memory-lifecycle.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/torchnpu-device-memory-lifecycle/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json
index d6ea1871c..ee8cdbfe6 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/ascend/skills/torchnpu-pinned-host-memory-lifecycle.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/SKILL.md",
"src/inferrsi/data/skill_library/hardware/ascend/torchnpu-pinned-host-memory-lifecycle/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json
index 12aeca1fd..1f6ba50d0 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-serving-preemption.json
@@ -16,8 +16,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/assess-serving-preemption/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/assess-serving-preemption/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/assess-serving-preemption/SKILL.md",
"src/inferrsi/data/skill_library/methods/assess-serving-preemption/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json
index 7c6e81473..b88ab7c72 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/assess-split-backward-schedule.json
@@ -16,8 +16,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/SKILL.md",
"src/inferrsi/data/skill_library/methods/assess-split-backward-schedule/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json
index a3b85ccdc..a3c4db1bf 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-fixed-workload-claim.json
@@ -20,8 +20,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/SKILL.md",
"src/inferrsi/data/skill_library/methods/audit-fixed-workload-claim/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json
index 50b90d31b..51509d4ec 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-metal-fasth3-semantics.json
@@ -29,8 +29,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/LICENSE",
"src/inferrsi/data/skill_library/repositories/h3-metal/audit-h3-metal-fasth3-semantics/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json
index c01e0e11f..b06208b30 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-h3-runtime-residency.json
@@ -30,8 +30,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/LICENSE",
"src/inferrsi/data/skill_library/repositories/h3-c/audit-h3-runtime-residency/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json
index 3adab00f6..6ee98cf8f 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-llama-runtime-state.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/SKILL.md",
"src/inferrsi/data/skill_library/repositories/llama-cpp/audit-llama-runtime-state/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json
index d5ab3d293..a1dbf5d14 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-recurrent-kv-finalization.json
@@ -26,8 +26,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/LICENSE",
"src/inferrsi/data/skill_library/repositories/vllm-rlt/audit-recurrent-kv-finalization/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json
index 8b883145a..72979e71e 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sd-cpp-residency.json
@@ -25,8 +25,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/LICENSE",
"src/inferrsi/data/skill_library/repositories/stable-diffusion-cpp/audit-sd-cpp-residency/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json
index 2dab44411..49672bc77 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-sglang-diffusion-lifetimes.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/LICENSE",
"src/inferrsi/data/skill_library/repositories/sglang/audit-sglang-diffusion-lifetimes/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json
index a4074b797..d2f6853da 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/audit-vllm-kv-lifetimes.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/SKILL.md",
"src/inferrsi/data/skill_library/repositories/vllm/audit-vllm-kv-lifetimes/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json
index 3c226dfe4..31adeb556 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/bound-windowed-attention-replay.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/SKILL.md",
"src/inferrsi/data/skill_library/repositories/deepinfer/bound-windowed-attention-replay/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json
index bb08f2ed5..1618acb9b 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/budget-speculative-verification.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/SKILL.md",
"src/inferrsi/data/skill_library/repositories/deepinfer/budget-speculative-verification/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json
index 3e86452fe..1f9de18d5 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/calibrate-serving-simulation.json
@@ -16,8 +16,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/SKILL.md",
"src/inferrsi/data/skill_library/methods/calibrate-serving-simulation/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json
index a88384b33..1651d5533 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cgroup-resource-controls.json
@@ -25,8 +25,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/SKILL.md",
"src/inferrsi/data/skill_library/hardware/host/cgroup-resource-controls/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json
index 3572a1f8c..ed8e689cf 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/checkpoint-recovery-consistency.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/SKILL.md",
"src/inferrsi/data/skill_library/methods/checkpoint-recovery-consistency/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json
index d5ce0bdbf..2584afe69 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/choose-parallel-training-plan.json
@@ -16,8 +16,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/SKILL.md",
"src/inferrsi/data/skill_library/methods/choose-parallel-training-plan/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json
index ded00b5e0..647907aa7 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/cite-check-downstream-pr.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/workflows/cite-check-downstream-pr/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/workflows/cite-check-downstream-pr/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/skill-index.json",
"src/inferrsi/data/skill_library/workflows/cite-check-downstream-pr/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json
index a7f2bfa5d..a00c666e7 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/claim-contract.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/workflows/claim-contract/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/workflows/claim-contract/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/skill-index.json",
"src/inferrsi/data/skill_library/workflows/claim-contract/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json
index acdc57b97..22f71296b 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/comfy-ui.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/comfy-ui/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/comfy-ui/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/comfy-ui/LICENSE",
"src/inferrsi/data/skill_library/repositories/comfy-ui/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json
index 0f5b0fb34..6d70b55e1 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/competitor-scan.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/workflows/competitor-scan/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/workflows/competitor-scan/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/skill-index.json",
"src/inferrsi/data/skill_library/workflows/competitor-scan/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json
index 064e2ca25..7b6971e04 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dcp-async-save-performance.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/SKILL.md",
"src/inferrsi/data/skill_library/frameworks/pytorch/dcp-async-save-performance/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json
index b20edd811..19b34cce7 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dense-performance-evidence.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/dense-performance-evidence/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/dense-performance-evidence/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/dense-performance-evidence/SKILL.md",
"src/inferrsi/data/skill_library/methods/dense-performance-evidence/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json
index 70fd8e8a5..de10678bf 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diagnose-vllm-admission-pressure.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/SKILL.md",
"src/inferrsi/data/skill_library/repositories/vllm/diagnose-vllm-admission-pressure/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json
index 3a883cf9a..fb71093f4 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/diffusers-invocation-state.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/SKILL.md",
"src/inferrsi/data/skill_library/frameworks/diffusers/diffusers-invocation-state/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json
index e90c2488d..70394982f 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/draft-acceptance-issue.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/workflows/draft-acceptance-issue/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/workflows/draft-acceptance-issue/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/skill-index.json",
"src/inferrsi/data/skill_library/workflows/draft-acceptance-issue/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json
index c456f30e9..e747e2e7a 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/dwarfstar.json
@@ -25,8 +25,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/dwarfstar/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/dwarfstar/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/dwarfstar/LICENSE",
"src/inferrsi/data/skill_library/repositories/dwarfstar/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json
index 3482d1338..3b5b4f37f 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/evaluate-elastic-serving-placement.json
@@ -16,8 +16,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/SKILL.md",
"src/inferrsi/data/skill_library/methods/evaluate-elastic-serving-placement/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json
index 7d68adf8b..f895c6f69 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/generation-cache-lifecycle.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/SKILL.md",
"src/inferrsi/data/skill_library/repositories/mlx-lm/generation-cache-lifecycle/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json
index bb39eac1d..fa89dffa5 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-memory-management.json
@@ -30,8 +30,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/host/host-memory-management/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/host/host-memory-management/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/host/host-memory-management/SKILL.md",
"src/inferrsi/data/skill_library/hardware/host/host-memory-management/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json
index 9f20ba9d7..61fcecd96 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/host-overhead-optimization.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/SKILL.md",
"src/inferrsi/data/skill_library/hardware/host/host-overhead-optimization/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json
index 3dee4b666..93cda492d 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/husky.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/husky/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/husky/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/husky/SKILL.md",
"src/inferrsi/data/skill_library/repositories/husky/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json
index 3b2327374..07f1b6628 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-benchmark-design.json
@@ -8,7 +8,7 @@
"component": "skills",
"scenario": "skills.discovery",
"status": "imported",
- "summary": "Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference. Covers AIPerf AgentX/Exgentic replay, vLLM/SGLang benchmark boundaries, SLO-aware metrics and InferRSI integration.",
+ "summary": "Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image, video and VLA inference. Covers AIPerf AgentX/Exgentic replay, vLLM/SGLang benchmark boundaries, VLA action-chunk timing and quality, SLO-aware metrics and InferRSI integration.",
"tags": [
"benchmark",
"evaluation",
@@ -27,7 +27,11 @@
"stream-boundaries",
"treatment-delivery",
"trial-validity",
- "protocol-audit"
+ "protocol-audit",
+ "vla",
+ "embodied",
+ "action-chunks",
+ "closed-loop"
],
"origin": [
"external"
@@ -35,8 +39,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json",
@@ -44,12 +48,14 @@
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/autotunebench-trial-validity.md",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/benchmark-commands.md",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/cache-aware-asr-stream-boundaries.md",
+ "src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-boundary-and-quality.md",
+ "src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-source.json",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/justgoal-evaluation-contracts.md",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/justgoal-source.json",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/task-benchmark-selection.md",
"src/inferrsi/data/skill_library/skill-index.json"
],
- "content_sha256": "7fd7f557f5c99ca4ec4c46f2d97e67ecf280f5f0fdf071871a14cc08a7cb1c84"
+ "content_sha256": "c4982a5e08dc89ad1f0db244325be796d29b9768e392c522b874facda47be694"
}
],
"index_type": "packaged-skill",
@@ -58,12 +64,14 @@
},
"checkout_ref": "checkout:inference-benchmark-design",
"source_files": {
- "src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md": "7fd7f557f5c99ca4ec4c46f2d97e67ecf280f5f0fdf071871a14cc08a7cb1c84",
- "src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json": "a646de55162cb8160f3f0947cab4b6ecac81927c3225019044b3f909087ba048",
+ "src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md": "c4982a5e08dc89ad1f0db244325be796d29b9768e392c522b874facda47be694",
+ "src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json": "5c9888da5b63f7ef08c3066fa22c1bbe39aafdd371566733cc1f54394138d5b7",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/autotunebench-source.json": "bc343ece2bf40c4e8aa4bd8d84f32465d00633255557feda5fe7799ce24e8688",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/autotunebench-trial-validity.md": "cb3a7a5d94d056f868af07f92a16a6fef31eafd3199ed07ae79e93798e320b1d",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/benchmark-commands.md": "0898424e0bff3a6246e1eeeda6abd6c5bd974c047fe32f3bbd5ab9d31ce17cd3",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/cache-aware-asr-stream-boundaries.md": "4e0ea9c1bf6b096a49a91ad5266e90b8a456f4f5aca15eba8ac270a53c2309b5",
+ "src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-boundary-and-quality.md": "0bc3ca87159cce960c7b32f62b82c275f850b98077ae07da0481bd554a3cd71d",
+ "src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-source.json": "056985be689554570a9f12b8e228ce71bdca5be940f9a2d7f6e1963e15a48130",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/justgoal-evaluation-contracts.md": "d25df795fd301e8a1bb436e5f8122d0bac2441c43104e59a3b310fd5a64f2f3e",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/justgoal-source.json": "60f2784384ea1a3b08a2ecafb60207d8e3707247dc7bc177b95b463d81f1cc16",
"src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/task-benchmark-selection.md": "0d4bb080a754e26846e81680023db0dff5e6a4c96be4327c411c3b72ac4ef8e2",
@@ -71,7 +79,7 @@
},
"skill_metadata": {
"name": "inference-benchmark-design",
- "description": "Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference. Covers AIPerf AgentX/Exgentic replay, vLLM/SGLang benchmark boundaries, SLO-aware metrics and InferRSI integration.",
+ "description": "Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image, video and VLA inference. Covers AIPerf AgentX/Exgentic replay, vLLM/SGLang benchmark boundaries, VLA action-chunk timing and quality, SLO-aware metrics and InferRSI integration.",
"tags": [
"benchmark",
"evaluation",
@@ -90,7 +98,11 @@
"stream-boundaries",
"treatment-delivery",
"trial-validity",
- "protocol-audit"
+ "protocol-audit",
+ "vla",
+ "embodied",
+ "action-chunks",
+ "closed-loop"
],
"status": "candidate",
"profile": "local-skill-v1",
@@ -310,14 +322,53 @@
"Tables 1-7"
],
"readset": "references/autotunebench-trial-validity.md"
+ },
+ {
+ "type": "git",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp",
+ "commit": "156bed151826780e2ce0132f8d0e3540889ae8cd",
+ "paths": [
+ "adapter/typed_io.py",
+ "eval/client/libero_profile.py",
+ "eval/client/vla_cpp_client.py",
+ "eval/conf/libero_smolvla_object_q8_eval.yaml",
+ "eval/README.md",
+ "eval/smolvla_q8_fixed_boundary_20261002.json",
+ "eval/smolvla_q8_object_followup_20260930.md",
+ "eval/SMOLVLA_VALIDATION.md",
+ "eval/THREE_MODEL_VALIDATION.md",
+ "eval/VLA_BENCHMARK_STANDARD.md",
+ "LICENSE",
+ "README.md",
+ "scripts/aggregate_smolvla_object_eval.py",
+ "scripts/bench_vla_boundary.py",
+ "scripts/profile_smolvla_python_baseline.py",
+ "serving/vla-server.cpp",
+ "serving/vla.proto"
+ ],
+ "readset": "references/embodied-cpp-source.json",
+ "license": "Apache-2.0",
+ "scope": "Selected CPU observation-to-action benchmark, client/profiler/aggregation and author artifact contracts; source-derived methodology, not native execution or quality/performance acceptance"
+ },
+ {
+ "type": "paper",
+ "id": "arxiv:2607.02501",
+ "version": "v3",
+ "url": "https://arxiv.org/abs/2607.02501v3",
+ "sections": [
+ "4 (VLA discussion and Table 3)"
+ ],
+ "readset": "references/embodied-cpp-source.json",
+ "license": "CC-BY-4.0",
+ "scope": "Independently normalized VLA metric semantics; predates the later SmolVLA experiment"
}
],
"empirical_validation": "not_run",
"applicability": {
- "scope": "Selecting task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference, and designing goodput-based multi-regime comparisons that fit InferRSI's evidence trail; interpreting treatment delivery, arm validity and stated-versus-observed protocol enforcement in supplied tuning reports",
- "prerequisites": "The primary objective, timing boundary, quality gates, sample/failure accounting, required regimes, repetitions, seeds and decision rule are frozen before optimization; baseline and candidate share hardware, software stack, input distribution, benchmark configuration, warmup and measurement protocol, and on a shared machine also one time window and dedicated devices; cache-aware ASR also requires the selected model/config and first/steady geometry, decoded clock and reset/EOU/EOF boundaries to be established",
- "input_contract": "The task family and deployment use case; candidate quality evaluators and reference parity; the declared SLO and the caller's policy on which regimes must improve or must not regress; for an audit, a supplied fixed-workload report with request-to-trial grouping, exclusion and failure counts and the declared aggregation rule; for a treatment audit, the frozen delivery/arm-validity rules, run/candidate identities, feedback and consumer-input evidence, and per-arm raw records",
- "produced_artifact": "A frozen benchmark design with evaluators, reference parity, primary objective, timing boundary, quality gates, required regimes, repetition and seed rules and an explicit decision config, plus the evidence-tree layout a PASS claim must produce; for a supplied report, the computed threshold result and missing evidence. Candidate knowledge, not qualification of any model or backend; for a treatment audit, a contract-bound eligibility and interpretation report separating treatment exposure, observed checks, comparator identity, anomaly disposition and source conflicts",
+ "scope": "Selecting task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference, and designing goodput-based multi-regime comparisons that fit InferRSI's evidence trail; interpreting treatment delivery, arm validity and stated-versus-observed protocol enforcement in supplied tuning reports; designing or auditing VLA deployment comparisons with explicit observation-to-action boundaries, generated/replayed horizons, per-task quality and source/precision identity",
+ "prerequisites": "The primary objective, timing boundary, quality gates, sample/failure accounting, required regimes, repetitions, seeds and decision rule are frozen before optimization; baseline and candidate share hardware, software stack, input distribution, benchmark configuration, warmup and measurement protocol, and on a shared machine also one time window and dedicated devices; cache-aware ASR also requires the selected model/config and first/steady geometry, decoded clock and reset/EOU/EOF boundaries to be established; VLA comparisons also freeze generated/replayed horizons, initial-state/noise protocol and intended precision or quantization treatment",
+ "input_contract": "The task family and deployment use case; candidate quality evaluators and reference parity; the declared SLO and the caller's policy on which regimes must improve or must not regress; for an audit, a supplied fixed-workload report with request-to-trial grouping, exclusion and failure counts and the declared aggregation rule; for a treatment audit, the frozen delivery/arm-validity rules, run/candidate identities, feedback and consumer-input evidence, and per-arm raw records; for VLA audits, the declared CPU input/output boundary, source/fixture/checkpoint/processor identities, per-task trial scope and available raw records or explicit artifact gaps",
+ "produced_artifact": "A frozen benchmark design with evaluators, reference parity, primary objective, timing boundary, quality gates, required regimes, repetition and seed rules and an explicit decision config, plus the evidence-tree layout a PASS claim must produce; for a supplied report, the computed threshold result and missing evidence. Candidate knowledge, not qualification of any model or backend; for a treatment audit, a contract-bound eligibility and interpretation report separating treatment exposure, observed checks, comparator identity, anomaly disposition and source conflicts; for VLA reports, a timing/denominator and precision reconciliation with per-task quality/completion accounting and source-artifact limits",
"failure_modes": [
"Raw throughput is optimized in place of goodput, so requests served outside the SLO count toward a win",
"One concurrency level or one undisclosed aggregate score stands in for all required regimes",
@@ -329,16 +380,22 @@
"A cache-aware ASR EOU is treated as EOF or full reset, nominal chunk duration as the decoded mel frontier, or shape-only geometry PASS as a native model-state test",
"An enabled treatment or stored feedback summary is taken as proof of delivery to the next eligible consumer",
"A failed arm-validity condition is overridden by a favorable best timing or a saturation/null interpretation",
- "Complete gate keys or zero recorded failures are read as proof that every declared protocol check ran"
+ "Complete gate keys or zero recorded failures are read as proof that every declared protocol check ran",
+ "Generated-action amortization is read as action-response latency or actual control frequency",
+ "Matching structural pair fields are taken to certify matched effective compute precision or quality equivalence",
+ "An aggregate VLA success rate conceals task-specific losses or missing/skipped episodes",
+ "A public current script, recorded experiment revision or binary hash is treated as the exact measured dirty source patch",
+ "Whole-device fallback memory is relabelled as process-attributed memory or compared with simulator-inclusive/allocator quantities"
],
- "refusal_point": "Records missing coverage and high-variance results as INCONCLUSIVE rather than a win; a run without a frozen RunSpec, or whose validator changed after the plan was frozen, is INCONCLUSIVE regardless of metric values; sets no universal numeric gate or noise threshold; withholds causal treatment benefit/null conclusions when required delivery or arm eligibility is unverified, and preserves conflicts between stated protocol, exported code and source reports"
+ "refusal_point": "Records missing coverage and high-variance results as INCONCLUSIVE rather than a win; a run without a frozen RunSpec, or whose validator changed after the plan was frozen, is INCONCLUSIVE regardless of metric values; sets no universal numeric gate or noise threshold; withholds causal treatment benefit/null conclusions when required delivery or arm eligibility is unverified, and preserves conflicts between stated protocol, exported code and source reports; withholds VLA quality-preservation, runtime-only attribution, control-rate or current-build conclusions when their required precision, per-task, timestamp or measured-source evidence is missing"
},
"recommendations": [
"Freeze one primary objective, timing boundary, quality gates and sample/failure accounting before optimization.",
"Choose the required regimes from the deployment use case, report each separately, and record which must improve or must not regress in a versioned decision config.",
"Use the replay trace that matches the actual deployment pattern.",
"State the concurrency, trace, model, SKU and microbenchmark scope boundaries in any report that cites benchmark results.",
- "Trace treatment to its eligible consumer and apply the frozen arm-validity precedence before attributing benefit, saturation or a null."
+ "Trace treatment to its eligible consumer and apply the frozen arm-validity precedence before attributing benefit, saturation or a null.",
+ "Separate complete VLA request cost, generated/replayed action denominators and actual control cadence; retain per-task quality and measured-source/precision limits."
],
"limitations": [
"Source-derived candidate; no local efficacy validation.",
@@ -347,7 +404,8 @@
"Benchmark feasibility and package-consumption checks do not promote this candidate or replace task quality, repeated A/B and service qualification.",
"The pinned NeMo-Speech.cpp ASR reference covers selected source/test definitions and author-reported benchmark boundaries; native model/device, decoder internals, quality, long-audio/load and service behavior remain not_run.",
"JustGoal contributes source-only controlled-evaluation and kernel-acceptance guidance; its repaired/retested timings, released-pass counts and protocol differences are author-reported, not locally reproduced or a complete campaign runner.",
- "The AutoTuneBench reference is source-only: original operational checkpoint/prompt receipts remain unavailable, source contradictions remain explicit, and no tuning, device/performance or agent-efficacy validation was run."
+ "The AutoTuneBench reference is source-only: original operational checkpoint/prompt receipts remain unavailable, source contradictions remain explicit, and no tuning, device/performance or agent-efficacy validation was run.",
+ "The Embodied.cpp reference is source-derived: its public latency arrays allow arithmetic audit, but raw episode recount, exact measured script/dirty patch, effective dispatch and current model/device/service or efficacy qualification remain unresolved."
],
"provenance_readsets": [
"references/task-benchmark-selection.md",
@@ -356,7 +414,9 @@
"references/justgoal-evaluation-contracts.md",
"references/justgoal-source.json",
"references/autotunebench-trial-validity.md",
- "references/autotunebench-source.json"
+ "references/autotunebench-source.json",
+ "references/embodied-cpp-boundary-and-quality.md",
+ "references/embodied-cpp-source.json"
],
"references": [
"../inference-optimization-casebook/cases/admission-drain-occupancy.md",
@@ -366,10 +426,10 @@
"searchspace_not_applicable": "SKILL.md §§ Select the task and its evaluation contract, Statistical requirements, Shared-machine measurement discipline, and Search integrity and final-state verification design comparable evaluation; hardware identity and allocation isolation are comparison controls, not a hardware-resource mechanism."
},
"applicability": {
- "scope": "Selecting task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference, and designing goodput-based multi-regime comparisons that fit InferRSI's evidence trail; interpreting treatment delivery, arm validity and stated-versus-observed protocol enforcement in supplied tuning reports",
- "prerequisites": "The primary objective, timing boundary, quality gates, sample/failure accounting, required regimes, repetitions, seeds and decision rule are frozen before optimization; baseline and candidate share hardware, software stack, input distribution, benchmark configuration, warmup and measurement protocol, and on a shared machine also one time window and dedicated devices; cache-aware ASR also requires the selected model/config and first/steady geometry, decoded clock and reset/EOU/EOF boundaries to be established",
- "input_contract": "The task family and deployment use case; candidate quality evaluators and reference parity; the declared SLO and the caller's policy on which regimes must improve or must not regress; for an audit, a supplied fixed-workload report with request-to-trial grouping, exclusion and failure counts and the declared aggregation rule; for a treatment audit, the frozen delivery/arm-validity rules, run/candidate identities, feedback and consumer-input evidence, and per-arm raw records",
- "produced_artifact": "A frozen benchmark design with evaluators, reference parity, primary objective, timing boundary, quality gates, required regimes, repetition and seed rules and an explicit decision config, plus the evidence-tree layout a PASS claim must produce; for a supplied report, the computed threshold result and missing evidence. Candidate knowledge, not qualification of any model or backend; for a treatment audit, a contract-bound eligibility and interpretation report separating treatment exposure, observed checks, comparator identity, anomaly disposition and source conflicts",
+ "scope": "Selecting task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference, and designing goodput-based multi-regime comparisons that fit InferRSI's evidence trail; interpreting treatment delivery, arm validity and stated-versus-observed protocol enforcement in supplied tuning reports; designing or auditing VLA deployment comparisons with explicit observation-to-action boundaries, generated/replayed horizons, per-task quality and source/precision identity",
+ "prerequisites": "The primary objective, timing boundary, quality gates, sample/failure accounting, required regimes, repetitions, seeds and decision rule are frozen before optimization; baseline and candidate share hardware, software stack, input distribution, benchmark configuration, warmup and measurement protocol, and on a shared machine also one time window and dedicated devices; cache-aware ASR also requires the selected model/config and first/steady geometry, decoded clock and reset/EOU/EOF boundaries to be established; VLA comparisons also freeze generated/replayed horizons, initial-state/noise protocol and intended precision or quantization treatment",
+ "input_contract": "The task family and deployment use case; candidate quality evaluators and reference parity; the declared SLO and the caller's policy on which regimes must improve or must not regress; for an audit, a supplied fixed-workload report with request-to-trial grouping, exclusion and failure counts and the declared aggregation rule; for a treatment audit, the frozen delivery/arm-validity rules, run/candidate identities, feedback and consumer-input evidence, and per-arm raw records; for VLA audits, the declared CPU input/output boundary, source/fixture/checkpoint/processor identities, per-task trial scope and available raw records or explicit artifact gaps",
+ "produced_artifact": "A frozen benchmark design with evaluators, reference parity, primary objective, timing boundary, quality gates, required regimes, repetition and seed rules and an explicit decision config, plus the evidence-tree layout a PASS claim must produce; for a supplied report, the computed threshold result and missing evidence. Candidate knowledge, not qualification of any model or backend; for a treatment audit, a contract-bound eligibility and interpretation report separating treatment exposure, observed checks, comparator identity, anomaly disposition and source conflicts; for VLA reports, a timing/denominator and precision reconciliation with per-task quality/completion accounting and source-artifact limits",
"failure_modes": [
"Raw throughput is optimized in place of goodput, so requests served outside the SLO count toward a win",
"One concurrency level or one undisclosed aggregate score stands in for all required regimes",
@@ -381,16 +441,22 @@
"A cache-aware ASR EOU is treated as EOF or full reset, nominal chunk duration as the decoded mel frontier, or shape-only geometry PASS as a native model-state test",
"An enabled treatment or stored feedback summary is taken as proof of delivery to the next eligible consumer",
"A failed arm-validity condition is overridden by a favorable best timing or a saturation/null interpretation",
- "Complete gate keys or zero recorded failures are read as proof that every declared protocol check ran"
+ "Complete gate keys or zero recorded failures are read as proof that every declared protocol check ran",
+ "Generated-action amortization is read as action-response latency or actual control frequency",
+ "Matching structural pair fields are taken to certify matched effective compute precision or quality equivalence",
+ "An aggregate VLA success rate conceals task-specific losses or missing/skipped episodes",
+ "A public current script, recorded experiment revision or binary hash is treated as the exact measured dirty source patch",
+ "Whole-device fallback memory is relabelled as process-attributed memory or compared with simulator-inclusive/allocator quantities"
],
- "refusal_point": "Records missing coverage and high-variance results as INCONCLUSIVE rather than a win; a run without a frozen RunSpec, or whose validator changed after the plan was frozen, is INCONCLUSIVE regardless of metric values; sets no universal numeric gate or noise threshold; withholds causal treatment benefit/null conclusions when required delivery or arm eligibility is unverified, and preserves conflicts between stated protocol, exported code and source reports"
+ "refusal_point": "Records missing coverage and high-variance results as INCONCLUSIVE rather than a win; a run without a frozen RunSpec, or whose validator changed after the plan was frozen, is INCONCLUSIVE regardless of metric values; sets no universal numeric gate or noise threshold; withholds causal treatment benefit/null conclusions when required delivery or arm eligibility is unverified, and preserves conflicts between stated protocol, exported code and source reports; withholds VLA quality-preservation, runtime-only attribution, control-rate or current-build conclusions when their required precision, per-task, timestamp or measured-source evidence is missing"
},
"recommendations": [
"Freeze one primary objective, timing boundary, quality gates and sample/failure accounting before optimization.",
"Choose the required regimes from the deployment use case, report each separately, and record which must improve or must not regress in a versioned decision config.",
"Use the replay trace that matches the actual deployment pattern.",
"State the concurrency, trace, model, SKU and microbenchmark scope boundaries in any report that cites benchmark results.",
- "Trace treatment to its eligible consumer and apply the frozen arm-validity precedence before attributing benefit, saturation or a null."
+ "Trace treatment to its eligible consumer and apply the frozen arm-validity precedence before attributing benefit, saturation or a null.",
+ "Separate complete VLA request cost, generated/replayed action denominators and actual control cadence; retain per-task quality and measured-source/precision limits."
],
"limitations": [
"This generated abstract points to the complete packaged skill; external evidence and upstream repositories remain source links.",
@@ -403,7 +469,8 @@
"Benchmark feasibility and package-consumption checks do not promote this candidate or replace task quality, repeated A/B and service qualification.",
"The pinned NeMo-Speech.cpp ASR reference covers selected source/test definitions and author-reported benchmark boundaries; native model/device, decoder internals, quality, long-audio/load and service behavior remain not_run.",
"JustGoal contributes source-only controlled-evaluation and kernel-acceptance guidance; its repaired/retested timings, released-pass counts and protocol differences are author-reported, not locally reproduced or a complete campaign runner.",
- "The AutoTuneBench reference is source-only: original operational checkpoint/prompt receipts remain unavailable, source contradictions remain explicit, and no tuning, device/performance or agent-efficacy validation was run."
+ "The AutoTuneBench reference is source-only: original operational checkpoint/prompt receipts remain unavailable, source contradictions remain explicit, and no tuning, device/performance or agent-efficacy validation was run.",
+ "The Embodied.cpp reference is source-derived: its public latency arrays allow arithmetic audit, but raw episode recount, exact measured script/dirty patch, effective dispatch and current model/device/service or efficacy qualification remain unresolved."
],
"verification": {
"locally_verified": false,
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json
index 3d6372c43..4fe0a4fff 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/inference-optimization-casebook.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/inference-optimization-casebook/SKILL.md",
"src/inferrsi/data/skill_library/methods/inference-optimization-casebook/cases/admission-drain-occupancy.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json
index 1040189df..b86689090 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kernel-writing-guide.json
@@ -39,8 +39,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/kernel-writing-guide/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/kernel-writing-guide/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/kernel-writing-guide/SKILL.md",
"src/inferrsi/data/skill_library/methods/kernel-writing-guide/cases/causal-conv1d-dispatch-state.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json
index 1c1f0ac80..f4611f505 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/kv-tier-management.json
@@ -24,8 +24,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/SKILL.md",
"src/inferrsi/data/skill_library/repositories/lmcache/kv-tier-management/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json
index c3b8b61d3..5a62693cc 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lazy-evaluation-stream-lifetimes.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/SKILL.md",
"src/inferrsi/data/skill_library/frameworks/mlx/lazy-evaluation-stream-lifetimes/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json
index 8c592e80d..4f25eea16 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lily.json
@@ -25,8 +25,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/lily/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/lily/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/lily/LICENSE",
"src/inferrsi/data/skill_library/repositories/lily/NOTICE.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json
index 26722f7f4..89e5c5db0 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/lithos-metal.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/lithos-metal/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/lithos-metal/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/lithos-metal/LICENSE",
"src/inferrsi/data/skill_library/repositories/lithos-metal/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json
index cb94c8a24..9467bc54b 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/local-dream.json
@@ -29,8 +29,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/local-dream/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/local-dream/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/local-dream/SKILL.md",
"src/inferrsi/data/skill_library/repositories/local-dream/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json
index 636fbe213..f24c48aef 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mistral-rs.json
@@ -26,8 +26,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/mistral-rs/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/mistral-rs/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/mistral-rs/LICENSE",
"src/inferrsi/data/skill_library/repositories/mistral-rs/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json
index b94c5f561..c32f02393 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlc-llm.json
@@ -27,8 +27,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/mlc-llm/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/mlc-llm/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/mlc-llm/LICENSE",
"src/inferrsi/data/skill_library/repositories/mlc-llm/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json
index fef01f452..ca4764686 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/mlx-swift-lm.json
@@ -27,8 +27,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/mlx-swift-lm/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/mlx-swift-lm/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/mlx-swift-lm/LICENSE",
"src/inferrsi/data/skill_library/repositories/mlx-swift-lm/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json
index 1adcbdc00..d2f725f2a 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/npu-skill-factory.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/workflows/npu-skill-factory/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/workflows/npu-skill-factory/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/skill-index.json",
"src/inferrsi/data/skill_library/workflows/npu-skill-factory/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json
index 8965fb642..e80223473 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/optimization-search-policy.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/optimization-search-policy/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/optimization-search-policy/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/optimization-search-policy/SKILL.md",
"src/inferrsi/data/skill_library/methods/optimization-search-policy/cases/sarathi-chunk-budget.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json
index 8c0cf7b12..932602afc 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/os-process-placement.json
@@ -24,8 +24,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/host/os-process-placement/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/host/os-process-placement/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/host/os-process-placement/SKILL.md",
"src/inferrsi/data/skill_library/hardware/host/os-process-placement/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json
index ae1810d82..9de6d0cda 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/performance-ceiling-analysis.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/SKILL.md",
"src/inferrsi/data/skill_library/methods/performance-ceiling-analysis/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json
index 5ccb28ffe..0489ed6a7 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/plan-activation-residency.json
@@ -16,8 +16,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/plan-activation-residency/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/plan-activation-residency/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/plan-activation-residency/SKILL.md",
"src/inferrsi/data/skill_library/methods/plan-activation-residency/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json
index abec7ae74..63b8b5e60 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/project-maya.json
@@ -25,8 +25,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/project-maya/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/project-maya/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/project-maya/LICENSE",
"src/inferrsi/data/skill_library/repositories/project-maya/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json
index 97d160bec..339971511 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/pytorch.json
@@ -24,8 +24,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/LICENSE",
"src/inferrsi/data/skill_library/frameworks/pytorch/pytorch/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json
index 27968f7c8..1a6034c0b 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/repository-evolution-analysis.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/repository-evolution-analysis/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/repository-evolution-analysis/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/repository-evolution-analysis/SKILL.md",
"src/inferrsi/data/skill_library/methods/repository-evolution-analysis/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json
index 465739baa..cab0f7674 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/rsi-evidence.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/workflows/rsi-evidence/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/workflows/rsi-evidence/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/skill-index.json",
"src/inferrsi/data/skill_library/workflows/rsi-evidence/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json
index c470e93b8..efa56ebe3 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/skill-evolution-playbook.json
@@ -21,8 +21,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/skill-evolution-playbook/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/skill-evolution-playbook/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/skill-evolution-playbook/SKILL.md",
"src/inferrsi/data/skill_library/methods/skill-evolution-playbook/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json
index a9476f7cf..cb2be7c3a 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/splash.json
@@ -25,8 +25,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/splash/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/splash/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/splash/LICENSE",
"src/inferrsi/data/skill_library/repositories/splash/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json
index fc715e09f..01230b40e 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/storage-io-measurement.json
@@ -24,8 +24,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/SKILL.md",
"src/inferrsi/data/skill_library/hardware/host/storage-io-measurement/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json
index 045b18bc4..b46cebe36 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-input-pipeline.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/training-input-pipeline/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/training-input-pipeline/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/training-input-pipeline/SKILL.md",
"src/inferrsi/data/skill_library/methods/training-input-pipeline/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json
index a2429cac2..4483126ea 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/training-resource-budget.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/training-resource-budget/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/training-resource-budget/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/training-resource-budget/SKILL.md",
"src/inferrsi/data/skill_library/methods/training-resource-budget/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json
index 2fbb6eebc..fa9f078f0 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/video-diffusion-optimization.json
@@ -26,8 +26,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/video-diffusion-optimization/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/video-diffusion-optimization/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/video-diffusion-optimization/SKILL.md",
"src/inferrsi/data/skill_library/methods/video-diffusion-optimization/assets/campaign.template.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json
index 179f342ea..e1e3bffb7 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/vllm-metal.json
@@ -26,8 +26,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/vllm-metal/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/vllm-metal/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/vllm-metal/LICENSE",
"src/inferrsi/data/skill_library/repositories/vllm-metal/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json
index 3cc50bdee..61944c413 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/common/skills/weight-loading-optimization.json
@@ -29,8 +29,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/methods/weight-loading-optimization/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/methods/weight-loading-optimization/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/methods/weight-loading-optimization/SKILL.md",
"src/inferrsi/data/skill_library/methods/weight-loading-optimization/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json
index 42b227ef5..6000947f4 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-afd-p2p-exchanges.json
@@ -23,8 +23,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/SKILL.md",
"src/inferrsi/data/skill_library/repositories/afd-plugin/audit-afd-p2p-exchanges/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json
index 0fc99822c..90b82d719 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/audit-flare-session-lifecycle.json
@@ -29,8 +29,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/LICENSE",
"src/inferrsi/data/skill_library/repositories/flare-local-inference/audit-flare-session-lifecycle/SKILL.md",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json
index 19f847d38..c529f6a78 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/diagnose-attention-bottlenecks.json
@@ -16,8 +16,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/SKILL.md",
"src/inferrsi/data/skill_library/hardware/cuda/diagnose-attention-bottlenecks/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json
index 7fac0bc38..da29d9002 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/host-overhead-profiling.json
@@ -22,8 +22,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/SKILL.md",
"src/inferrsi/data/skill_library/hardware/cuda/host-overhead-profiling/metadata.json",
diff --git a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json
index 55040913f..3655f89f1 100644
--- a/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json
+++ b/src/inferrsi/data/repo_packs/inferrsi/knowledge/cuda/skills/reason-hopper-kernels.json
@@ -36,8 +36,8 @@
"sources": [
{
"type": "git",
- "url": "https://github.com/InferMatrixLab/InferRSI/blob/7aa3344bf84c42467f43b3a927e1c7d13642bd55/src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/SKILL.md",
- "commit": "7aa3344bf84c42467f43b3a927e1c7d13642bd55",
+ "url": "https://github.com/InferMatrixLab/InferRSI/blob/d5fc2121d6d1628111403ba70ec3b70638ab8674/src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/SKILL.md",
+ "commit": "d5fc2121d6d1628111403ba70ec3b70638ab8674",
"paths": [
"src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/SKILL.md",
"src/inferrsi/data/skill_library/hardware/cuda/reason-hopper-kernels/metadata.json",
diff --git a/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md
index b40a52d2c..5bc72a8f0 100644
--- a/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md
+++ b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/SKILL.md
@@ -1,6 +1,6 @@
---
name: inference-benchmark-design
-description: 'Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference. Covers AIPerf AgentX/Exgentic replay, vLLM/SGLang benchmark boundaries, SLO-aware metrics and InferRSI integration.'
+description: 'Select task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image, video and VLA inference. Covers AIPerf AgentX/Exgentic replay, vLLM/SGLang benchmark boundaries, VLA action-chunk timing and quality, SLO-aware metrics and InferRSI integration.'
---
# Inference Benchmark Design
@@ -24,6 +24,13 @@ before selecting tests or comparing chunk latency. It separates startup/steady g
progress, stream-slot reuse, reset, EOU and EOF; its source tests and
author benchmark reports are not local model or serving qualification.
+For VLA and robot deployment comparisons, read
+[observation-to-action timing and closed-loop quality](references/embodied-cpp-boundary-and-quality.md).
+Define the complete CPU-action boundary, distinguish generated and replayed
+horizons from the control clock, and reconcile per-task quality, precision and
+the measured source revision. Freeze latency/jitter and quality objectives for
+the actual control use case; per-generated-action cost is amortization.
+
For treatment-enabled tuning comparisons or claimed protocol integrity, read
[treatment delivery and trial validity](references/autotunebench-trial-validity.md).
Trace feedback to the eligible consumer input and apply the frozen arm-validity
diff --git a/src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json
index e6ea1ff3b..e15840c2a 100644
--- a/src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json
+++ b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/metadata.json
@@ -17,7 +17,11 @@
"stream-boundaries",
"treatment-delivery",
"trial-validity",
- "protocol-audit"
+ "protocol-audit",
+ "vla",
+ "embodied",
+ "action-chunks",
+ "closed-loop"
],
"status": "candidate",
"profile": "local-skill-v1",
@@ -237,14 +241,53 @@
"Tables 1-7"
],
"readset": "references/autotunebench-trial-validity.md"
+ },
+ {
+ "type": "git",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp",
+ "commit": "156bed151826780e2ce0132f8d0e3540889ae8cd",
+ "paths": [
+ "adapter/typed_io.py",
+ "eval/client/libero_profile.py",
+ "eval/client/vla_cpp_client.py",
+ "eval/conf/libero_smolvla_object_q8_eval.yaml",
+ "eval/README.md",
+ "eval/smolvla_q8_fixed_boundary_20261002.json",
+ "eval/smolvla_q8_object_followup_20260930.md",
+ "eval/SMOLVLA_VALIDATION.md",
+ "eval/THREE_MODEL_VALIDATION.md",
+ "eval/VLA_BENCHMARK_STANDARD.md",
+ "LICENSE",
+ "README.md",
+ "scripts/aggregate_smolvla_object_eval.py",
+ "scripts/bench_vla_boundary.py",
+ "scripts/profile_smolvla_python_baseline.py",
+ "serving/vla-server.cpp",
+ "serving/vla.proto"
+ ],
+ "readset": "references/embodied-cpp-source.json",
+ "license": "Apache-2.0",
+ "scope": "Selected CPU observation-to-action benchmark, client/profiler/aggregation and author artifact contracts; source-derived methodology, not native execution or quality/performance acceptance"
+ },
+ {
+ "type": "paper",
+ "id": "arxiv:2607.02501",
+ "version": "v3",
+ "url": "https://arxiv.org/abs/2607.02501v3",
+ "sections": [
+ "4 (VLA discussion and Table 3)"
+ ],
+ "readset": "references/embodied-cpp-source.json",
+ "license": "CC-BY-4.0",
+ "scope": "Independently normalized VLA metric semantics; predates the later SmolVLA experiment"
}
],
"empirical_validation": "not_run",
"applicability": {
- "scope": "Selecting task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference, and designing goodput-based multi-regime comparisons that fit InferRSI's evidence trail; interpreting treatment delivery, arm validity and stated-versus-observed protocol enforcement in supplied tuning reports",
- "prerequisites": "The primary objective, timing boundary, quality gates, sample/failure accounting, required regimes, repetitions, seeds and decision rule are frozen before optimization; baseline and candidate share hardware, software stack, input distribution, benchmark configuration, warmup and measurement protocol, and on a shared machine also one time window and dedicated devices; cache-aware ASR also requires the selected model/config and first/steady geometry, decoded clock and reset/EOU/EOF boundaries to be established",
- "input_contract": "The task family and deployment use case; candidate quality evaluators and reference parity; the declared SLO and the caller's policy on which regimes must improve or must not regress; for an audit, a supplied fixed-workload report with request-to-trial grouping, exclusion and failure counts and the declared aggregation rule; for a treatment audit, the frozen delivery/arm-validity rules, run/candidate identities, feedback and consumer-input evidence, and per-arm raw records",
- "produced_artifact": "A frozen benchmark design with evaluators, reference parity, primary objective, timing boundary, quality gates, required regimes, repetition and seed rules and an explicit decision config, plus the evidence-tree layout a PASS claim must produce; for a supplied report, the computed threshold result and missing evidence. Candidate knowledge, not qualification of any model or backend; for a treatment audit, a contract-bound eligibility and interpretation report separating treatment exposure, observed checks, comparator identity, anomaly disposition and source conflicts",
+ "scope": "Selecting task quality gates, reference parity checks and performance workloads for text, agentic coding, work agents, cache-aware ASR, TTS, full-duplex audio/video, image and video inference, and designing goodput-based multi-regime comparisons that fit InferRSI's evidence trail; interpreting treatment delivery, arm validity and stated-versus-observed protocol enforcement in supplied tuning reports; designing or auditing VLA deployment comparisons with explicit observation-to-action boundaries, generated/replayed horizons, per-task quality and source/precision identity",
+ "prerequisites": "The primary objective, timing boundary, quality gates, sample/failure accounting, required regimes, repetitions, seeds and decision rule are frozen before optimization; baseline and candidate share hardware, software stack, input distribution, benchmark configuration, warmup and measurement protocol, and on a shared machine also one time window and dedicated devices; cache-aware ASR also requires the selected model/config and first/steady geometry, decoded clock and reset/EOU/EOF boundaries to be established; VLA comparisons also freeze generated/replayed horizons, initial-state/noise protocol and intended precision or quantization treatment",
+ "input_contract": "The task family and deployment use case; candidate quality evaluators and reference parity; the declared SLO and the caller's policy on which regimes must improve or must not regress; for an audit, a supplied fixed-workload report with request-to-trial grouping, exclusion and failure counts and the declared aggregation rule; for a treatment audit, the frozen delivery/arm-validity rules, run/candidate identities, feedback and consumer-input evidence, and per-arm raw records; for VLA audits, the declared CPU input/output boundary, source/fixture/checkpoint/processor identities, per-task trial scope and available raw records or explicit artifact gaps",
+ "produced_artifact": "A frozen benchmark design with evaluators, reference parity, primary objective, timing boundary, quality gates, required regimes, repetition and seed rules and an explicit decision config, plus the evidence-tree layout a PASS claim must produce; for a supplied report, the computed threshold result and missing evidence. Candidate knowledge, not qualification of any model or backend; for a treatment audit, a contract-bound eligibility and interpretation report separating treatment exposure, observed checks, comparator identity, anomaly disposition and source conflicts; for VLA reports, a timing/denominator and precision reconciliation with per-task quality/completion accounting and source-artifact limits",
"failure_modes": [
"Raw throughput is optimized in place of goodput, so requests served outside the SLO count toward a win",
"One concurrency level or one undisclosed aggregate score stands in for all required regimes",
@@ -256,16 +299,22 @@
"A cache-aware ASR EOU is treated as EOF or full reset, nominal chunk duration as the decoded mel frontier, or shape-only geometry PASS as a native model-state test",
"An enabled treatment or stored feedback summary is taken as proof of delivery to the next eligible consumer",
"A failed arm-validity condition is overridden by a favorable best timing or a saturation/null interpretation",
- "Complete gate keys or zero recorded failures are read as proof that every declared protocol check ran"
+ "Complete gate keys or zero recorded failures are read as proof that every declared protocol check ran",
+ "Generated-action amortization is read as action-response latency or actual control frequency",
+ "Matching structural pair fields are taken to certify matched effective compute precision or quality equivalence",
+ "An aggregate VLA success rate conceals task-specific losses or missing/skipped episodes",
+ "A public current script, recorded experiment revision or binary hash is treated as the exact measured dirty source patch",
+ "Whole-device fallback memory is relabelled as process-attributed memory or compared with simulator-inclusive/allocator quantities"
],
- "refusal_point": "Records missing coverage and high-variance results as INCONCLUSIVE rather than a win; a run without a frozen RunSpec, or whose validator changed after the plan was frozen, is INCONCLUSIVE regardless of metric values; sets no universal numeric gate or noise threshold; withholds causal treatment benefit/null conclusions when required delivery or arm eligibility is unverified, and preserves conflicts between stated protocol, exported code and source reports"
+ "refusal_point": "Records missing coverage and high-variance results as INCONCLUSIVE rather than a win; a run without a frozen RunSpec, or whose validator changed after the plan was frozen, is INCONCLUSIVE regardless of metric values; sets no universal numeric gate or noise threshold; withholds causal treatment benefit/null conclusions when required delivery or arm eligibility is unverified, and preserves conflicts between stated protocol, exported code and source reports; withholds VLA quality-preservation, runtime-only attribution, control-rate or current-build conclusions when their required precision, per-task, timestamp or measured-source evidence is missing"
},
"recommendations": [
"Freeze one primary objective, timing boundary, quality gates and sample/failure accounting before optimization.",
"Choose the required regimes from the deployment use case, report each separately, and record which must improve or must not regress in a versioned decision config.",
"Use the replay trace that matches the actual deployment pattern.",
"State the concurrency, trace, model, SKU and microbenchmark scope boundaries in any report that cites benchmark results.",
- "Trace treatment to its eligible consumer and apply the frozen arm-validity precedence before attributing benefit, saturation or a null."
+ "Trace treatment to its eligible consumer and apply the frozen arm-validity precedence before attributing benefit, saturation or a null.",
+ "Separate complete VLA request cost, generated/replayed action denominators and actual control cadence; retain per-task quality and measured-source/precision limits."
],
"limitations": [
"Source-derived candidate; no local efficacy validation.",
@@ -274,7 +323,8 @@
"Benchmark feasibility and package-consumption checks do not promote this candidate or replace task quality, repeated A/B and service qualification.",
"The pinned NeMo-Speech.cpp ASR reference covers selected source/test definitions and author-reported benchmark boundaries; native model/device, decoder internals, quality, long-audio/load and service behavior remain not_run.",
"JustGoal contributes source-only controlled-evaluation and kernel-acceptance guidance; its repaired/retested timings, released-pass counts and protocol differences are author-reported, not locally reproduced or a complete campaign runner.",
- "The AutoTuneBench reference is source-only: original operational checkpoint/prompt receipts remain unavailable, source contradictions remain explicit, and no tuning, device/performance or agent-efficacy validation was run."
+ "The AutoTuneBench reference is source-only: original operational checkpoint/prompt receipts remain unavailable, source contradictions remain explicit, and no tuning, device/performance or agent-efficacy validation was run.",
+ "The Embodied.cpp reference is source-derived: its public latency arrays allow arithmetic audit, but raw episode recount, exact measured script/dirty patch, effective dispatch and current model/device/service or efficacy qualification remain unresolved."
],
"provenance_readsets": [
"references/task-benchmark-selection.md",
@@ -283,7 +333,9 @@
"references/justgoal-evaluation-contracts.md",
"references/justgoal-source.json",
"references/autotunebench-trial-validity.md",
- "references/autotunebench-source.json"
+ "references/autotunebench-source.json",
+ "references/embodied-cpp-boundary-and-quality.md",
+ "references/embodied-cpp-source.json"
],
"references": [
"../inference-optimization-casebook/cases/admission-drain-occupancy.md",
diff --git a/src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-boundary-and-quality.md b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-boundary-and-quality.md
new file mode 100644
index 000000000..7c077627d
--- /dev/null
+++ b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-boundary-and-quality.md
@@ -0,0 +1,220 @@
+# VLA benchmark boundaries and closed-loop quality
+
+Use this procedure for vision-language-action (VLA) deployment comparisons,
+latency-per-action or control-frequency claims, and quantized runtime reports.
+The [source record](embodied-cpp-source.json) retains the pinned Embodied.cpp
+repository, companion paper v3, selected readset, rights and evidence limits.
+
+## Freeze the observation-to-action contract
+
+Name the input available when timing starts and the output required when it
+ends. Record image layout/resolution, state dimensions, instruction, noise,
+batch size and output shape. Trace preprocessing, tokenization, host/device
+transfers, synchronization, postprocessing and CPU output materialization on
+both paths. Include the deployed transport and state any asymmetry. Separate
+model loading, fixture construction, sensor/simulator work and action replay
+from that timer.
+
+The selected fixed benchmark takes two CHW 512-pixel CPU images, raw 8-D state,
+an exact instruction and a 50x32 noise array, and returns a complete 50x7 CPU
+action chunk. Its Python path calls the official processors, model and
+postprocessor before CPU materialization. Its C++ path includes client
+preprocessing, local ZMQ request/response and conversion of the returned chunk;
+the server serializes a CPU float vector. This boundary excludes simulator
+execution and sensor acquisition. The separate closed-loop experiment has
+360-pixel observations and 512-pixel model images; do not silently give the
+fixed-fixture test that larger input boundary.
+
+Freeze the campaign's latency/jitter objective, quality gates, repetitions,
+seeds and failure accounting before comparison. The source's five warmups,
+100 timed requests and twenty separate untimed memory requests describe its
+protocol, not universal InferRSI defaults.
+
+Sources: [benchmark predictors and timed loop][bench], [client][client],
+[server response][server] and [benchmark standard][standard].
+
+## Separate generated actions, replay and control clocks
+
+For request i producing H_i actions, L_i / H_i is cost per generated action.
+It does not say when an action becomes available to a robot or how often a
+fresh observation is used. Record the number of actions replayed before
+replanning and measure actual observation/action timestamps separately.
+
+| Quantity | Numerator and denominator | Meaning |
+|---|---|---|
+| Complete request latency | CPU observation to complete CPU action chunk | The declared deployment request |
+| Fixed-boundary cost per generated action | Complete request latency / generated horizon | Amortized generation cost |
+| Source profiler inf_ms | Server-reported latency_ms_total per unique request | A narrower server timing scope |
+| Profiler generated_action_step_ms | Server-reported time / generated chunk size | Server-scope generation amortization |
+| Profiler model_step_ms | Server-reported time / configured replay count | Replay-amortized server cost |
+| Control cadence | Actual action/observation timestamps, queueing and replay | The execution clock |
+
+Equal denominators do not make server-only and complete client timing
+comparable. The source client generates a chunk, queues only the configured
+replay prefix and returns one action at a time. The Object experiment uses
+generation horizon 50 and replay horizon 1. About 308 ms/request divided by
+50 gives 6.161 ms/generated action; it does not establish 6.161 ms action
+response latency or roughly 162 Hz control.
+
+Sources: [client queue/request methods][client], [profiler metrics][profiler]
+and the standard's performance protocol.
+
+## Reconcile identity and the intended treatment
+
+Compare source checkpoint, tokenizer/processor settings, normalization and
+per-instruction layouts; fixture, instruction and noise identities; RNG
+device/dtype; flow settings; output shape, generated horizon and replay horizon;
+hardware allocation and software stack. Equal tensor names do not establish
+equal inference configuration.
+
+Keep source checkpoint storage, converted/resident weight storage and
+effective compute precision distinct. A runtime-only comparison needs matched
+effective precision. If quantization or compute policy is the intended
+difference, name the joint runtime/precision treatment and assess its quality
+trade-off rather than attributing the entire difference to runtime overhead.
+
+The inspected pair gate checks backend pairing, plan, boundary/script/fixture/
+instruction/checkpoint/tokenizer identities, shape, horizon, memory source and
+GPU identity. It does not compare effective compute precision or establish
+output-quality equivalence. The retained Python manifest reports BF16/F32
+parameter dtypes and disabled autocast; C++ reports Q8_0 resident weights and
+F32 compute. These are author manifests, not proof of each executed operator's
+dtype.
+
+Sources: [pair validation and result manifests][bench], [standard][standard]
+and [retained artifact][artifact].
+
+## Audit per-task quality and the complete trial denominator
+
+Freeze benchmark version, suite/tasks, episodes per task, initial-state order,
+control mode, image/prompt construction, noise/seed protocol, action replay,
+generated horizon and success definition. Keep requested, evaluable,
+successful, skipped and aborted trials visible. Recount task/episode identities
+from raw records before making a complete-quality claim. Report per-task
+counts alongside totals and the declared uncertainty intervals.
+
+The source's strict Object aggregation checks task/episode IDs and protocol,
+cross-checks profile/task records, rejects skips, aborts and incomplete raw
+samples, and reconciles quantizer/startup manifests. A complete single-arm
+result still sets promotion_eligible=false because the matching comparator is
+separate. Those are static source checks until execution and inputs are
+established.
+
+The Q8 follow-up reports C++ 166/200 versus Python 172/200 across ten Object
+tasks with twenty episodes each. Task 7 is 9/20 versus 20/20. Preserve that
+loss when interpreting the aggregate 83% versus 86%. Overlapping Wilson
+intervals do not establish equivalence or noninferiority; use the campaign's
+frozen quality policy. Keep this Object comparison separate from historical
+four-suite runs under an older seed protocol and later small smoke samples.
+
+Sources: [strict aggregation][aggregate], [follow-up][followup],
+[historical validation][validation] and [ledger][ledger].
+
+## Interpret the retained sample arrays and memory scope
+
+The committed [artifact][artifact] contains 100 latency samples per backend.
+Recomputing those arrays produces the values below. This is an audit of
+published samples, not a fresh device measurement.
+
+| Observation | Python | C++ Q8_0 | Boundary |
+|---|---:|---:|---|
+| Request mean | 490.755 ms | 308.030 ms | Complete fixed CPU-input-to-CPU-output request |
+| Request p99 | 518.767 ms | 393.719 ms | Same timing boundary and sample arrays |
+| Mean / generated horizon 50 | 9.815 ms | 6.161 ms | Amortized generation cost |
+| Sampled peak memory | 1671 MiB | 1029 MiB | Isolated whole-device WSL fallback |
+
+The request-latency ratio is approximately 0.627665; the sampled device-memory
+ratio is 0.615799. Both memory records declare process_level_vram=false.
+Twenty untimed memory requests produced 73 Python and 13 C++ sampler values:
+periodic observations, not twenty per-request memory measurements. Retain the
+measurement source and author allocation-isolation claim. Do not relabel
+whole-device memory as process-attributed memory or compare allocator bytes,
+file size and simulator-inclusive memory as identical quantities.
+
+The closed-loop Python profiler times inside _get_action_chunk with CUDA
+synchronization; its memory includes the simulator. Those diagnostic values
+are not the denominator for the fixed-boundary table. The follow-up identifies
+one timing interval interrupted by host-process suspension using an exact
+task/index/value and reason, with raw data and filtered/unfiltered summaries
+retained. That disposition does not authorize threshold-based outlier removal
+or excluding unsuccessful trials from a future comparison.
+
+Sources: [artifact][artifact], [follow-up][followup], [memory sampler][profiler]
+and [Python profiler/aggregation][baseline].
+
+## Bind measurements to the code actually measured
+
+Keep inspection revision, declared experiment revision, actual script/binary
+identities and the complete dirty patch separate.
+
+| Identity | Retained observation |
+|---|---|
+| Inspected repository | 156bed151826780e2ce0132f8d0e3540889ae8cd |
+| Current pinned benchmark script SHA-256 | 2a7e94a4df662d2da41169cc2345b75e683c82e082fe590eba713302255e1a7b |
+| Artifact's declared experiment revision | 3a3468cfe634e88d93d08aafcae89bd3ef0fb5a2 |
+| Public script at that revision SHA-256 | 3ab7f108eb4493dec548e8641a65e54d3dd605bab27076e17abf1422792f9332; its architecture choices omit SmolVLA |
+| Recorded measured-script SHA-256 | dee5573089d882ec5fcade6c99ede546ab810c22526c894c68bbca471175a892 |
+| Closed-loop source base | dcedfec5721768661bd81f4196a0325d55f814f9 plus uncommitted changes |
+
+Neither public script identity matches the recorded measured script. The
+follow-up retains a server-binary hash, but a binary hash or dirty-path count
+does not identify the complete source patch. Request exact measured
+source/patch and artifact identities before reproduction or current-build
+attribution; code presence cannot transfer historical speed or quality.
+
+The public tree retains compact latency arrays and author quality tables.
+Referenced raw episode logs, the fixture and complete measurement patch are
+absent. When episode recount is unavailable, retain the author-reported
+counts and withhold the affected quality/reproduction claim. Missing artifacts
+do not erase available sample arithmetic or justify invented replacement data.
+
+Sources: [follow-up provenance][followup], [artifact][artifact],
+[current script][bench] and [public historical script][historical].
+
+## Return a comparison at its supported evidence level
+
+Produce a timing/denominator table, input/precision reconciliation, per-task
+quality and completion account, artifact/source identities and unresolved
+receipts. Distinguish a runtime-only matched comparison from a named joint
+runtime/precision trade-off. Withhold quality-preservation, current-build or
+control-rate conclusions when their required observations are missing.
+
+Name the missing evidence needed to recover: original episode IDs/outcomes,
+the measured script/dirty patch, matched precision/dispatch or actual
+observation/action timestamps. Future execution needs its own authorized,
+frozen campaign. This reference supplies methodology and source interpretation,
+not a runtime integration.
+
+## Attribution and source scope
+
+This is original analysis of SEU-PAISys/Embodied.cpp and Ling Xu et al.,
+*Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on
+Heterogeneous Robots*, [arXiv:2607.02501v3][paper], dated 2026-08-09.
+The paper is [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/);
+the repository declares [Apache-2.0][license], retaining Copyright 2026
+VinRobotics for original portions and Copyright 2026 SEU-PAISys for
+modifications and new files. These notes adapt methodology in original prose;
+paper figures/full text, source code and raw arrays are not redistributed.
+Dependencies, checkpoints and datasets retain separate rights.
+
+Paper v3/Table 3 supplies independently normalized VLA latency, success and
+memory semantics. It predates the September/October SmolVLA follow-up.
+Broader WAM/runtime-support, model math and quantizer correctness are outside
+this readset's qualification. This source-derived candidate establishes no
+native dispatch, model/device correctness, quality preservation, service
+performance or agent efficacy.
+
+[bench]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/scripts/bench_vla_boundary.py
+[client]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/client/vla_cpp_client.py
+[server]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/serving/vla-server.cpp
+[standard]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/VLA_BENCHMARK_STANDARD.md
+[profiler]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/client/libero_profile.py
+[aggregate]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/scripts/aggregate_smolvla_object_eval.py
+[followup]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/smolvla_q8_object_followup_20260930.md
+[artifact]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/smolvla_q8_fixed_boundary_20261002.json
+[baseline]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/scripts/profile_smolvla_python_baseline.py
+[validation]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/SMOLVLA_VALIDATION.md
+[ledger]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/THREE_MODEL_VALIDATION.md
+[historical]: https://github.com/SEU-PAISys/Embodied.cpp/blob/3a3468cfe634e88d93d08aafcae89bd3ef0fb5a2/scripts/bench_vla_boundary.py
+[paper]: https://arxiv.org/abs/2607.02501v3
+[license]: https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/LICENSE
diff --git a/src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-source.json b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-source.json
new file mode 100644
index 000000000..f5ed57a8e
--- /dev/null
+++ b/src/inferrsi/data/skill_library/methods/inference-benchmark-design/references/embodied-cpp-source.json
@@ -0,0 +1,455 @@
+{
+ "title": "Embodied.cpp VLA deployment benchmark source record",
+ "repository": {
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp",
+ "commit": "156bed151826780e2ce0132f8d0e3540889ae8cd",
+ "tree": "cff35978852e8199e176cee2eba0322b7af3eb90",
+ "tree_entries": 213,
+ "blob_entries": 188,
+ "tree_truncated": false,
+ "default_license_declaration": "Apache-2.0"
+ },
+ "inspected_on": "2026-11-04",
+ "paired_paper": {
+ "id": "arxiv:2607.02501",
+ "version": "v3",
+ "url": "https://arxiv.org/abs/2607.02501v3",
+ "html_url": "https://arxiv.org/html/2607.02501v3",
+ "title": "Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots",
+ "authors": "Ling Xu et al.",
+ "version_date": "2026-08-09",
+ "license": "CC-BY-4.0",
+ "license_url": "https://creativecommons.org/licenses/by/4.0/",
+ "read_scope": "Version/license/project-link metadata; selected architecture context in 3.2; section 4 including Tables 3-4 and conclusion. Only normalized VLA metric semantics from section 4/Table 3 are proposed for reuse. No paper figures, PDF, full text or WAM qualification is proposed for redistribution.",
+ "reuse_scope": "Section 4/Table 3 independently normalized VLA metric semantics. The later SmolVLA follow-up is a separate experiment, not a paper-v3 result."
+ },
+ "rights": {
+ "repository_license_sha256": "88993458addec64e6922f552eb27ddd5f775d911db56205f167c885800011bf7",
+ "attribution": [
+ "Copyright 2026 SEU-PAISys",
+ "Original portions Copyright 2026 VinRobotics",
+ "Modifications and new files Copyright 2026 SEU-PAISys"
+ ],
+ "file_notice_scope": "Selected client/server/protobuf/typed-I/O and profiler notices explicitly declare Apache-2.0; selected benchmark/aggregation/baseline scripts have no separate file notice and use the repository default declaration",
+ "separate_rights": "README line 594 reserves third-party dependencies, checkpoints, datasets and upstream implementations to their own licenses. No such artifacts are proposed for redistribution.",
+ "distribution": "Original attributed procedure prose and source links only; no upstream code, raw sample arrays, paper figures/full text, model weights or datasets are redistributed."
+ },
+ "files": [
+ {
+ "path": "adapter/typed_io.py",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/adapter/typed_io.py",
+ "bytes": 1656,
+ "git_blob_sha1": "7f20a7848fcbd8d1fa659f440fd2f9a433d6e3e0",
+ "sha256": "4b68fa78ad474f5dd31bb61c5c167114a4c04eff3d08f83d4e87d5a7cab0bdb1",
+ "read_scope": "Complete typed I/O boundary inspected as context; fixed benchmark directly builds model-input mappings, so no claim that this typed adapter executes in the benchmark",
+ "read_regions_one_based": [
+ [
+ 1,
+ 62
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/client/libero_profile.py",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/client/libero_profile.py",
+ "bytes": 21870,
+ "git_blob_sha1": "0ad704b61930937b781b2e454ebe044b7e328ba5",
+ "sha256": "340d208302f4c04218b40850eda31159eba0a5de4368bdb457c81339ffee7c6e",
+ "read_scope": "Selected memory sampling and complete metric/episode accumulation/result methods; report-formatting helpers excluded",
+ "read_regions_one_based": [
+ [
+ 1,
+ 34
+ ],
+ [
+ 64,
+ 140
+ ],
+ [
+ 204,
+ 501
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/client/vla_cpp_client.py",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/client/vla_cpp_client.py",
+ "bytes": 37607,
+ "git_blob_sha1": "e68c4a9339e4a5c29f33061e1f90317acbc28ab5",
+ "sha256": "bfc875aded705c4b50965b2d50f492c19997372188e84ba3b4655b1a5325aab9",
+ "read_scope": "Selected complete reachable SmolVLA setup/predict/request/queue/close methods; unrelated XR0/X-VLA prediction bodies excluded",
+ "read_regions_one_based": [
+ [
+ 1,
+ 134
+ ],
+ [
+ 203,
+ 226
+ ],
+ [
+ 369,
+ 742
+ ],
+ [
+ 931,
+ 941
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/conf/libero_smolvla_object_q8_eval.yaml",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/conf/libero_smolvla_object_q8_eval.yaml",
+ "bytes": 1062,
+ "git_blob_sha1": "ef8d8716f725c4d08a6e4417b56e92c39eb4cf4f",
+ "sha256": "7ca892ee8dfeaebd16b87cc0338d86145e2f6bab632e53a3e2ef92686b413bd6",
+ "read_scope": "Complete Q8 Object control/episode/image/replay/generated-horizon configuration",
+ "read_regions_one_based": [
+ [
+ 1,
+ 50
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/README.md",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/README.md",
+ "bytes": 6388,
+ "git_blob_sha1": "ca4190cb62b51abed6550fa00d10fe3db6e88da5",
+ "sha256": "22a77e97436acc788903c4a40bbe5ea90a47431823ec2f69c46bfb9124353ebe",
+ "read_scope": "Complete evaluation navigation and historical/smoke/acceptance distinctions",
+ "read_regions_one_based": [
+ [
+ 1,
+ 137
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/smolvla_q8_fixed_boundary_20261002.json",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/smolvla_q8_fixed_boundary_20261002.json",
+ "bytes": 18632,
+ "git_blob_sha1": "2b9498f51964dcbfaa9d97ff7e28eed9f7251cc5",
+ "sha256": "db53513cfdc52a03c3a4f2f1a2bfbab7a1eed45a9995873ac3bf34b7e9d7019d",
+ "read_scope": "Complete JSON structure; both 100-request arrays independently recomputed; precision/source-version/memory metadata and aggregate success claims inspected",
+ "read_regions_one_based": [],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/smolvla_q8_object_followup_20260930.md",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/smolvla_q8_object_followup_20260930.md",
+ "bytes": 10658,
+ "git_blob_sha1": "36013b63f3c9f589754edbedecc7401ac081e442",
+ "sha256": "db8691f6e745424e2466d98e7611b50c7080ff95d782d2ee8fdb78d0537862fd",
+ "read_scope": "Complete Q8 follow-up; fixed-boundary results, per-task quality, historical diagnostics, exclusions, dirty-source and raw-artifact limits",
+ "read_regions_one_based": [
+ [
+ 1,
+ 228
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/SMOLVLA_VALIDATION.md",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/SMOLVLA_VALIDATION.md",
+ "bytes": 8487,
+ "git_blob_sha1": "42dd49aa6048900e48baa857ded41a8f7e9b9959",
+ "sha256": "3dd8080c07529a25691e48c22800734d20fccfe4571d409c677396e1f96c107d",
+ "read_scope": "Complete historical SmolVLA validation; old seed protocol, smoke, numerical and unequal-timing-scope cautions",
+ "read_regions_one_based": [
+ [
+ 1,
+ 176
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/THREE_MODEL_VALIDATION.md",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/THREE_MODEL_VALIDATION.md",
+ "bytes": 5238,
+ "git_blob_sha1": "cd8115df57aaded4df2f11bcc0d718bca2488948",
+ "sha256": "d19faff9e7b70add7ca44d9e642c3783e2084454949e38e4ee29b3e55f7abe7b",
+ "read_scope": "Complete validation ledger; published versus reproduced versus pending and explicit precision/protocol/coverage gaps",
+ "read_regions_one_based": [
+ [
+ 1,
+ 61
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "eval/VLA_BENCHMARK_STANDARD.md",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/eval/VLA_BENCHMARK_STANDARD.md",
+ "bytes": 6984,
+ "git_blob_sha1": "e6758c911fee427958c8fb8ad82ad6bc6f82824f",
+ "sha256": "aa8ba3bd6982a63be4f8533973fc2b8d9ee4b68a4312dee788fa33851a343ce6",
+ "read_scope": "Complete protocol, timing inclusions/exclusions, metric denominators, quality/evidence rules",
+ "read_regions_one_based": [
+ [
+ 1,
+ 129
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "LICENSE",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/LICENSE",
+ "bytes": 11462,
+ "git_blob_sha1": "b45820ee42df0db9253634546b1c321343446137",
+ "sha256": "88993458addec64e6922f552eb27ddd5f775d911db56205f167c885800011bf7",
+ "read_scope": "Apache-2.0 identity, grants, redistribution conditions and VinRobotics/SEU-PAISys attribution",
+ "read_regions_one_based": [
+ [
+ 1,
+ 15
+ ],
+ [
+ 66,
+ 123
+ ],
+ [
+ 189,
+ 205
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "README.md",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/README.md",
+ "bytes": 25578,
+ "git_blob_sha1": "0479552cf71976ad5c365e6ff1e4181d0093ee9d",
+ "sha256": "60ec56863978989382aea8f20a0336da6a8cbc24c34162d96c81eb245f738198",
+ "read_scope": "Performance-source classification, SmolVLA replay/prompt caveats, source rights and third-party boundary",
+ "read_regions_one_based": [
+ [
+ 7,
+ 8
+ ],
+ [
+ 160,
+ 182
+ ],
+ [
+ 418,
+ 446
+ ],
+ [
+ 571,
+ 613
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "scripts/aggregate_smolvla_object_eval.py",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/scripts/aggregate_smolvla_object_eval.py",
+ "bytes": 41255,
+ "git_blob_sha1": "768e77510903c11fdbaa90d3953ca201d128bff7",
+ "sha256": "578dc3ed0c2cc814b5535294c0a5b801c8d602685656ced6f50f32bfe0d9894c",
+ "read_scope": "Complete static strict-aggregation path; manifests, task/episode IDs, skip rejection, raw sample checks and promotion_eligible=false",
+ "read_regions_one_based": [
+ [
+ 1,
+ 928
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "scripts/bench_vla_boundary.py",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/scripts/bench_vla_boundary.py",
+ "bytes": 27881,
+ "git_blob_sha1": "9a858e199b857106fe4a48dd09d05e793f2375ec",
+ "sha256": "2a7e94a4df662d2da41169cc2345b75e683c82e082fe590eba713302255e1a7b",
+ "read_scope": "Complete fixed-boundary benchmark; input preparation, Python/C++ timed predictors, pair gate, sampling and artifact output",
+ "read_regions_one_based": [
+ [
+ 1,
+ 511
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "scripts/profile_smolvla_python_baseline.py",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/scripts/profile_smolvla_python_baseline.py",
+ "bytes": 31128,
+ "git_blob_sha1": "68627bff2c1956620b3f2f09ccb0aea23a17fde6",
+ "sha256": "7216e578d692ac2a78c0d3b45f93e01a5c9501e7c039fbbf27b030a4cc0d13e8",
+ "read_scope": "Complete static Python baseline path; model-internal timing, seeds/resume, exact-index exclusion and filtered/unfiltered summaries",
+ "read_regions_one_based": [
+ [
+ 1,
+ 778
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "serving/vla-server.cpp",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/serving/vla-server.cpp",
+ "bytes": 25761,
+ "git_blob_sha1": "8dd8d681cf90dbb27a33002c8f6e8a19a9b24ce1",
+ "sha256": "d9d459cf966d320adea01e3835b88add55fb1d32a548c32eea1b478cdf69c7dc",
+ "read_scope": "Attribution plus model-input build, prediction result, complete CPU-vector response serialization and timing fields; other-model parse paths and model math excluded",
+ "read_regions_one_based": [
+ [
+ 1,
+ 39
+ ],
+ [
+ 460,
+ 554
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ },
+ {
+ "path": "serving/vla.proto",
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/156bed151826780e2ce0132f8d0e3540889ae8cd/serving/vla.proto",
+ "bytes": 2191,
+ "git_blob_sha1": "e15f623bce6592a88d8693e56bd7ff9387fcd0f7",
+ "sha256": "7cdc70febe06183b5c42aa272c811c8d23d811334406c56c3459457724745c70",
+ "read_scope": "Complete request/response boundary and attribution",
+ "read_regions_one_based": [
+ [
+ 1,
+ 79
+ ]
+ ],
+ "license": "Apache-2.0 (repository default; selected explicit file notices agree)"
+ }
+ ],
+ "measurement_evidence": {
+ "artifact_path": "eval/smolvla_q8_fixed_boundary_20261002.json",
+ "artifact_sha256": "db53513cfdc52a03c3a4f2f1a2bfbab7a1eed45a9995873ac3bf34b7e9d7019d",
+ "author_measurement_date": "2026-10-02",
+ "inspection_is_not_measurement": true,
+ "boundary_id": "raw_cpu_observation_to_complete_cpu_action_chunk.v1",
+ "fixed_input": {
+ "images": "two CPU float CHW images, 512x512 each",
+ "raw_state_dimensions": 8,
+ "noise_shape": [
+ 50,
+ 32
+ ],
+ "instruction": "Exact instruction bound by the artifact hash"
+ },
+ "output_shape": [
+ 50,
+ 7
+ ],
+ "generated_horizon": 50,
+ "measurement_plan": {
+ "n": 100,
+ "warmup": 5,
+ "memory_requests": 20,
+ "formal": true
+ },
+ "conditions": {
+ "hardware": "NVIDIA GeForce RTX 4060 Laptop GPU, sm_89, WSL2",
+ "cuda_driver": "CUDA 12.0 / driver 576.88",
+ "python_stack": "PyTorch 2.5.1 CUDA 12.4; LeRobot 0.4.3; LIBERO 0.1.0",
+ "quality_scope": "LIBERO-Object, tasks 0-9, 20 episodes per task per backend",
+ "quality_images": "360x360 observations / 512x512 model images",
+ "environment_seed": 1000,
+ "action_noise_seed": 1000,
+ "noise_protocol": "Derived per episode",
+ "replayed_actions_per_request": 1,
+ "flow_steps": 10
+ },
+ "reported_manifests": {
+ "python": {
+ "source_checkpoint_storage": "safetensors",
+ "runtime_parameter_dtypes": [
+ "torch.bfloat16",
+ "torch.float32"
+ ],
+ "autocast": false
+ },
+ "cpp_q8_0": {
+ "source_checkpoint_storage": "safetensors",
+ "gguf_resident_weight_storage": "Q8_0",
+ "runtime_compute": "F32"
+ },
+ "interpretation": "Different runtime/precision policies are a joint treatment; manifests are not independently verified operator dispatch."
+ },
+ "published_sample_summary": {
+ "python_request_count": 100,
+ "cpp_request_count": 100,
+ "python_request_mean_ms": 490.75542880999984,
+ "cpp_request_mean_ms": 308.0299474900005,
+ "python_request_p99_ms": 518.7674982100198,
+ "cpp_request_p99_ms": 393.7192080099521,
+ "python_generated_action_mean_ms": 9.815108576199998,
+ "cpp_generated_action_mean_ms": 6.1605989498000095,
+ "normalized_to_python": {
+ "latency_per_request": 0.6276648803191476,
+ "latency_per_generated_action": 0.6276648803191475,
+ "vram_peak": 0.6157989228007181
+ },
+ "interpretation": "Arithmetic over public author sample arrays; not a fresh device measurement or action-response/control-frequency result."
+ },
+ "memory": {
+ "source": "device_total_fallback",
+ "process_level_vram": false,
+ "python_sample_count": 73,
+ "cpp_sample_count": 13,
+ "python_sampled_peak_mib": 1671,
+ "cpp_sampled_peak_mib": 1029,
+ "interpretation": "Periodic whole-device samples on an author-described isolated GPU; not process-attributed or one sample per request."
+ },
+ "quality": {
+ "evidence": "Author aggregate and task tables; raw episode artifacts absent from the pinned tree",
+ "python_successes": 172,
+ "cpp_successes": 166,
+ "episodes_per_backend": 200,
+ "task_7": {
+ "python_successes": 20,
+ "cpp_successes": 9,
+ "episodes_per_backend": 20
+ },
+ "interpretation": "Retain per-task loss. Aggregate closeness and overlapping Wilson intervals do not establish equivalence or noninferiority."
+ },
+ "reported_exclusion": {
+ "scope": "Historical closed-loop Python timing diagnostics, not the fixed-boundary arrays",
+ "task_id": 1,
+ "sample_index": 2274,
+ "request_ms": 180864.500452,
+ "reason": "User-requested host process suspension",
+ "handling": "Exact-index exclusion described with raw sample retained and filtered/unfiltered summaries; no quality or memory sample removed"
+ }
+ },
+ "source_version_boundaries": {
+ "inspected_commit": "156bed151826780e2ce0132f8d0e3540889ae8cd",
+ "current_script_sha256": "2a7e94a4df662d2da41169cc2345b75e683c82e082fe590eba713302255e1a7b",
+ "artifact_declared_revision": "3a3468cfe634e88d93d08aafcae89bd3ef0fb5a2",
+ "public_script_at_declared_revision": {
+ "url": "https://github.com/SEU-PAISys/Embodied.cpp/blob/3a3468cfe634e88d93d08aafcae89bd3ef0fb5a2/scripts/bench_vla_boundary.py",
+ "sha256": "3ab7f108eb4493dec548e8641a65e54d3dd605bab27076e17abf1422792f9332",
+ "observation": "Public architecture choices omit SmolVLA"
+ },
+ "recorded_measured_script_sha256": "dee5573089d882ec5fcade6c99ede546ab810c22526c894c68bbca471175a892",
+ "closed_loop_base": "dcedfec5721768661bd81f4196a0325d55f814f9",
+ "closed_loop_tree_state": "Uncommitted changes; full source patch not supplied",
+ "reported_server_binary_sha256": "e45398c0bf2cd069a166e504a309683535988cc9f888e4f41257959034981210",
+ "interpretation": "Neither public script hash matches the measured-script hash; binary identity or dirty-path count does not identify the missing source patch."
+ },
+ "exclusions_and_limits": [
+ "Selected file regions support benchmark methodology, not whole-repository or WAM/runtime-support qualification.",
+ "Fixed fixtures, raw episode logs, complete measurement patch, weights and datasets are absent or excluded; no replacement data is invented.",
+ "Broader native kernels, model math, quantizer correctness and effective dtype dispatch are outside this qualification.",
+ "Source code, published sample arithmetic and author quality reports remain distinct from model/device/service and agent-efficacy validation.",
+ "Source numeric protocol values are contextual, not InferRSI default thresholds."
+ ]
+}