Skip to content

feat: improve ADR command consequences quality#188

Closed
tractorjuice wants to merge 1 commit intomainfrom
feat/improve-adr-prompt
Closed

feat: improve ADR command consequences quality#188
tractorjuice wants to merge 1 commit intomainfrom
feat/improve-adr-prompt

Conversation

@tractorjuice
Copy link
Copy Markdown
Owner

Summary

Strengthens the Consequences section of the ADR command prompt to produce more actionable output:

  • Project-specific metrics — require measurable outcomes tied to actual NFR/BR IDs (e.g., "API response time: 500ms → < 200ms per NFR-P-001") instead of generic "improved performance"
  • Requirement traceability — tie each positive consequence to a specific requirement or stakeholder goal
  • Explicit trade-offs — require mitigation strategies with named owners and timelines, not vague statements
  • After-action review — schedule a review 1 month after implementation to verify assumptions

How this was found

Used the autoresearch command prompt optimiser against the UK Gov ADR Framework and MADR v4.0 references:

Iteration Score Status Change
0 (baseline) 8.6 keep
1 9.0 keep Strengthen consequences with project-specific metrics and after-action review
2 9.2 discard Tie monitoring metrics to consequences (delta < 0.3)

Net: +3 lines of prompt, score 8.6 → 9.0

Test plan

  • Run /arckit:adr against a test project and verify consequences reference specific NFR/BR IDs
  • Verify trade-offs include named owners and timelines
  • Verify after-action review is scheduled in output
  • Run python scripts/converter.py to regenerate extension formats

🤖 Generated with Claude Code

Strengthen Consequences section to require project-specific
metrics tied to requirements, explicit trade-off mitigations
with owners and timelines, and after-action review schedule.

Found via autoresearch experiment loop (score 8.6 → 9.0).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
tractorjuice added a commit that referenced this pull request Mar 27, 2026
Merged 6 pre-autoresearch prompt quality PRs into the combined branch:

- #187 requirements: user personas, UC structure, Given/When/Then AC
- #188 adr: measurable consequences with project-specific metrics
- #189 backlog: strict acceptance criteria rules (testable, thresholds)
- #191 wardley: math-validated stage boundaries table
- #192 wardley.value-chain: visibility rationale and validation checklist
- #194 wardley.climate: recommended action field with specificity rules

All autoresearch additions (scores, traceability) preserved alongside
the quality improvements. Lint passes. Converter regenerated.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
@tractorjuice
Copy link
Copy Markdown
Owner Author

Merged into #265 (combined autoresearch PR)

@tractorjuice tractorjuice deleted the feat/improve-adr-prompt branch March 27, 2026 23:52
tractorjuice added a commit that referenced this pull request Apr 6, 2026
Merged 6 pre-autoresearch prompt quality PRs into the combined branch:

- #187 requirements: user personas, UC structure, Given/When/Then AC
- #188 adr: measurable consequences with project-specific metrics
- #189 backlog: strict acceptance criteria rules (testable, thresholds)
- #191 wardley: math-validated stage boundaries table
- #192 wardley.value-chain: visibility rationale and validation checklist
- #194 wardley.climate: recommended action field with specificity rules

All autoresearch additions (scores, traceability) preserved alongside
the quality improvements. Lint passes. Converter regenerated.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant