feat: knowledge·tooling·experts hub — architecture + per-repo generation recipes#10
Merged
Conversation
…ion-recipe layer Deep-dive on making the repo a hub that offers, per cited repo, a knowledge graph + agentic tooling — engineered high-quality/fast/cheap/fresh/future-proof. Core decision (docs/HUB_ARCHITECTURE.md): INDEX + JIT, not a warehouse of frozen artifacts. Keep a cheap always-fresh index committed; generate the expensive deep KG/skills just-in-time by REUSING existing engines (graphify / understand-anything / anyagent reverse), cached by commit SHA, exposed via MCP. Never pay for an unused repo; never ship a stale one. Phase-0 (this PR): the RECIPE layer — scripts/repos.py turns data/stack.yml + frontier into docs/REPOS.md (22 repos, 20 on GitHub), each with its KG + reverse command + provenance. Deterministic, gated in make check. Deliberately generates NO stale artifacts. E23. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Answers "for each cited repo, offer a knowledge graph + agentic tooling — done high-quality, fast, cheap, up-to-date, future-proof."
The deep dive (
docs/HUB_ARCHITECTURE.md): the deciding call is INDEX + JIT, not a warehouse of frozen artifacts. Eagerly minting a KG + skill + plugin + workflow for all ~27 repos fails every bar (slop, slow, costly, stale-next-day, unmaintainable). Instead, three layers:graphify/understand-anything/anyagent reverse, served via MCP.Plus the third pillar (experts) is already wired via
people.yml+ the field graph (repo → lab → expert).Phase-0 shipped (
scripts/repos.py→docs/REPOS.md): turnsstack.yml+ frontier into a per-repo generation-recipe index — 22 repos (20 GitHub), each with its KG engine +anyagent reversecommand + provenance. Deterministic, gated inmake check, zero wasteful generation. README hub pointer added; E23; audit 100/100, 12/23 executions.I deliberately did not generate 27 knowledge graphs + skills now — that's the anti-pattern the architecture exists to avoid. Deep generation is Phase 2, JIT + SHA-cached.
🤖 Generated with Claude Code