Port adr-writing and context-gathering skills from audio-processor#3
Closed
marquesds wants to merge 1 commit into
Closed
Port adr-writing and context-gathering skills from audio-processor#3marquesds wants to merge 1 commit into
marquesds wants to merge 1 commit into
Conversation
…ssor audio-processor's .agents/ harness had two generic skills that were missing from the upstream style harness: - adr-writing: SMART + STAR ADR format for architecture decisions that outlive a single PR. Generalized from the audio-processor version — service-specific subsystems and examples replaced with generic ones. - context-gathering: gather repo/Linear/Datadog/Notion/Slack/Git evidence before asking questions during requirements work. ai-contribution-disclosure was also considered but excluded — it was removed from style in a prior commit (c1cb68d) and the audio-processor copy is identical to what was deleted, so re-adding it would revert that decision. Catalog rows and eval cases added for both skills. Lint and tests pass.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Why
python/services/audio-processor/.agents/ships two generic skills that were missing from the upstream style harness. They were written there first, proved useful, and belong in the single source of truth so every agent install gets them — not just the audio-processor service.What changed
adr-writing— SMART + STAR ADR format for architecture decisions that outlive a single PR. Generalized from the audio-processor version: service-specific subsystems (audio-processor,sessions,transcription,chunking,websocket-runtime) and examples replaced with generic ones. Condensed the body template to stay under the 200-line file limit.context-gathering— gather repo/Linear/Datadog/Notion/Slack/Git evidence before asking questions during requirements work. Ported as-is with style frontmatter.skills-catalog.md— added 2 catalog rows (lint enforces catalog↔skills parity).skill-routing.yml— added 2 eval cases asserting the new skills load correctly.What was excluded
ai-contribution-disclosure— was also present in audio-processor but not ported. It was removed from style in commitc1cb68d("chore(skills): remove ai ack disclosure"), and the audio-processor copy is identical to what was deleted. Re-adding it would revert that decision without new rationale.audio-processor-package-placement,scribe-session-debug,orchestrator-parity— service-specific, no upstream value.Validation
uv run python -m scripts.lint_source→ 94 source files cleanuv run pytest→ 80 passed