AGENT SKILL

speckit-archive-readiness

Archive historical Spec Kit readiness evidence into specs/archive, leaving only a README stub and refreshed archive inventory. Use before a squash-merge commit is finalized, after a feature is no longer current gate evidence, or when the user asks to archive unnecessary readiness/evidence files.

License 许可
MIT
最近更新
2026-06-16
Stars
★ 2
Forks
0

01 前言 Front Matter

name
speckit-archive-readiness
metadata
{ "short-description": "Archive historical readiness evidence" }
description
Archive historical Spec Kit readiness evidence into specs/archive, leaving only a README stub and refreshed archive inventory. Use before a squash-merge commit is finalized, after a feature is no longer current gate evidence, or when the user asks to archive unnecessary readiness/evidence files.

03 SKILL.md 正文

speckit-archive-readiness

Archive historical feature readiness artifacts so old logs, transcripts, JSON hits, screenshots, and generated pass/fail exhaust stop polluting active checkout search while remaining available as audit context.

What Counts As Unnecessary

Archive only specs/<feature>/readiness/** for a feature that is no longer the current gate target. After a feature is squash-merged, its readiness files are historical audit context and must not be cited as current pass/fail evidence.

Do not archive:

  • the active feature before its merge commit is being finalized;
  • root readiness/**;
  • specs/archive/**;
  • protected archive-governance feature 036-archive-readiness-api-docs;
  • package API reference material unless the user explicitly asks.

Command

Use the deterministic script:

dotnet fsi scripts/archive-readiness.fsx --feature <feature-id>

For multiple historical features:

dotnet fsi scripts/archive-readiness.fsx --feature 065-typed-controls-front-door --feature 066-typed-catalog-generation

For review before writing:

dotnet fsi scripts/archive-readiness.fsx --dry-run --feature <feature-id>

The script writes:

  • specs/archive/<feature>/readiness.zip
  • specs/archive/<feature>/README.md
  • specs/<feature>/readiness/README.md
  • specs/archive/readiness-archives.json
  • refreshed 036-archive-readiness-api-docs archive inventory/current evidence map/stale-reference scan.

Merge Hook

During /speckit-merge, run this after git merge --squash <branch> and before the squash commit:

FEATURES=$(git diff --cached --name-only | sed -nE 's#^specs/([0-9]{3}-[^/]+)/readiness/.*#\1#p' | sort -u)
for FEATURE in $FEATURES; do
  dotnet fsi scripts/archive-readiness.fsx --feature "$FEATURE"
  git add -A "specs/$FEATURE/readiness" "specs/archive/$FEATURE" \
    "specs/archive/readiness-archives.json" \
    "specs/036-archive-readiness-api-docs/readiness/archive-inventory.md" \
    "specs/036-archive-readiness-api-docs/readiness/current-evidence-map.md" \
    "specs/036-archive-readiness-api-docs/readiness/stale-reference-scan.md" \
    "specs/036-archive-readiness-api-docs/readiness/stale-reference-scan.json"
done

If the script fails, stop the merge before committing. Do not delete a feature branch until the squash commit, archive conversion, and normal merge checks have all completed.

Restore

Historical evidence remains recoverable:

python3 -m zipfile -e specs/archive/<feature>/readiness.zip .

04 同仓库更多 Skills