Epic (PUBLIC)
Parent for product behaviour and platform infrastructure that make AI-assisted insights available in Telemetry Tracker.
This epic lives in the public repository. Assume everything here becomes public documentation.
Goal
Help developers understand what changed, what is affected, and what to investigate next — using telemetry-grounded summaries — without exposing how the private AI service is implemented.
In scope (public)
- Async workspace brief architecture (API, jobs, worker process boundaries)
- Durable persistence for completed briefs and generation jobs
- Acknowledgement and stale-brief semantics
- Dashboard brief / intelligence UI surfaces
- Public API contracts for requesting and reading summaries
- Worker deployment requirements for self-hosters
- Retention policies for brief/job data
- Observability of public API and workers (success/failure, queue depth — no model/prompt details)
- Self-hosting documentation for enabling the AI worker integration point
Out of scope (private repo)
Do not track here:
- Prompts, prompt versions, or model names
- Provider SDKs or Structured Outputs wiring
- Evaluation, hallucination, or factual-consistency heuristics
- Benchmarks, cost monitoring thresholds, or rollout/allowlist logic
- Ranking / significance algorithms or injection mitigations
Those belong in telemetry-tracker-ai under milestone v0.4.x — AI Intelligence (epic: AI Generation Pipeline).
Related public work
| Item |
Role |
| #496 |
Release / incident intelligence product surface |
| v1.10.x — Async workspace brief |
Shipped async brief infrastructure |
| #453 / v1.16.x |
Release Health (dependency, not AI implementation) |
| #196 #197 #495 / v1.17.x |
Performance / compare (dependency) |
Planning rule
Create child issues only after scope is clearly public product/infra. Child titles/bodies must describe behaviour for contributors, self-hosters, and users — never private LLM implementation.
Base branch
develop
Epic (PUBLIC)
Parent for product behaviour and platform infrastructure that make AI-assisted insights available in Telemetry Tracker.
This epic lives in the public repository. Assume everything here becomes public documentation.
Goal
Help developers understand what changed, what is affected, and what to investigate next — using telemetry-grounded summaries — without exposing how the private AI service is implemented.
In scope (public)
Out of scope (private repo)
Do not track here:
Those belong in
telemetry-tracker-aiunder milestone v0.4.x — AI Intelligence (epic: AI Generation Pipeline).Related public work
Planning rule
Create child issues only after scope is clearly public product/infra. Child titles/bodies must describe behaviour for contributors, self-hosters, and users — never private LLM implementation.
Base branch
develop