| name | feature-radar-archive |
| description | Archive a completed, rejected, or covered feature into .feature-radar/archive/ with mandatory
learning extraction. MUST use this skill whenever a feature reaches a terminal state — done,
rejected, covered, deferred, or N/A. Even casual mentions like "we shipped X" or "X is done"
should trigger this. The skill extracts learnings, derives new opportunities, and updates refs.
Use when the user:
- Says "we shipped X", "X is done", "X is complete", "we just finished X"
- Rejects a feature: "we decided not to build X", "reject this", "doesn't fit"
- Defers: "defer X", "postpone this", "revisit later", "not now"
- Closes an opportunity: "close this opportunity", "mark as done", "archive this"
- Mentions any feature reaching Done/Covered/Rejected/Deferred status
Do NOT use for discovering new features — that's feature-radar-scan's job.
|
Archive Feature
Move a feature to .feature-radar/archive/ and run the mandatory extraction checklist.
Deep Read
Read and follow `../feature-radar/references/DEEP-READ.md` — complete all 6 steps before proceeding.
Behavioral Directives
Read and follow `../feature-radar/references/DIRECTIVES.md`.
Workflow
-
Identify the feature — ask the user which feature to archive. Check opportunities/ for an existing file to move.
-
Determine status:
- Done — fully implemented and working
- Covered — existing functionality handles the use case
- Rejected — decided against implementing
- N/A — not applicable to our architecture
- Deferred — valuable but postponed with rationale
-
Create archive file — write .feature-radar/archive/{nn}-{slug}.md
-
Checkpoint — Review & Annotate per ../feature-radar/references/WORKFLOW-PATTERNS.md
-
Run extraction checklist:
Extraction Checklist
Complete ALL 5 checks below. For each check, you MUST explicitly state the finding.
Do NOT proceed to step 5 until every check has a written response.
□ archive/{nn}-{slug}.md created with correct status
□ Extract learnings → specs/{topic}.md
□ Derive new opportunities → opportunities/{nn}-{slug}.md
□ Update references → references/{topic}.md
□ Update ecosystem trends → specs/ecosystem-trends.md
Acceptable responses per check:
- "No learnings to extract" — acceptable, but must be stated
- "New opportunity identified: {description}" — create the file
- "No reference updates needed" — acceptable, but must be stated
- Remove from opportunities — if the feature had an
opportunities/ file, delete it
- Update base.md — adjust counts in Tracking Summary
Checkpoint: Present extraction results using this format:
Archive: {nn}-{slug} ({status})
| Check | Result |
|-------|--------|
| Learnings | {created specs/X.md / updated specs/X.md / none} |
| Opportunities | {created opportunities/X.md / none} |
| References | {updated references/X.md / none} |
| Ecosystem | {updated specs/ecosystem-trends.md / none} |
Archive File Format
Use the format defined in ../feature-radar/references/SPEC.md § 3.2 (archive/{nn}-{slug}.md).
Guidelines
- The extraction checklist is NOT optional. Every archive action must go through all 5 checks.
- If the feature was in
opportunities/, use the same number. If it's new, use the next available number.
- "Deferred" is not a trash bin — include a clear rationale and conditions for re-evaluation.
Example Output
→ Moved opportunities/04-config-merge.md → archive/04-config-merge.md (Done)
→ Learnings: specs/yaml-config-merge.md (new)
→ New opportunity: opportunities/09-config-validation.md
→ References: no updates | Ecosystem: no changes
→ Updated base.md: archive 3→4, specs 2→3
Completion Summary
Follow the template in ../feature-radar/references/DIRECTIVES.md, with skill name "Archive Complete".
