G
github-repo-candidate-quality-gate
About github-repo-candidate-quality-gate
GitHub Repo Candidate Quality Gate is a skill for filtering broad GitHub search results into recommendation-grade repository candidates. It retains discovery counts while applying metadata, freshness, fit, and traction checks, with lane-level status and degraded-cause reporting.
Key features
- 01Filter by license and archive status
- 02Apply repository freshness and fit checks
- 03Require traction or adoption rationale
- 04Preserve broad discovery counts
- 05Report lane status and degraded causes
Use cases
- 01Turn noisy GitHub searches into shortlist candidates
- 02Review repository recommendations for merge readiness
- 03Report search-lane health in cron reviews
Install
Copy the command and run it in a terminal.
$
npx skills add https:/ / github.com/ akillness/ jeo-skills --skill github-repo-candidate-quality-gateAfter installing, type/github-repo-candidate-quality-gatein Claude Code to confirm it loaded.
Bundled filesevals/evals.jsonSKILL.toon
Details
From the project READMEGitHub Repo Candidate Quality Gate
Use this skill when broad GitHub search output is too noisy for merge-ready recommendations.
When to use
- You already ran
gh search repos(or equivalent) and need recommendation-grade keeps. - You must preserve broad discovery evidence while filtering keeps deterministically.
- You need lane-level status + degraded-cause reporting for unattended cron review.
Minimum keep metadata
For every kept candidate, capture all fields:
full_nameurllicensepushed_at(orupdated_atfallback)archivedstars- one-line fit rationale
- provenance label
Quality gate order
- License gate: drop missing/unknown license unless explicit exception rationale.
- Archive gate: drop archived repos by default.
- Freshness gate: drop if
pushed_atolder than 24 months unless exception + risk. - Fit gate: drop keyword-only mismatches (low-fit).
- Signal gate: keep broad discovery evidence, but recommendation-grade keeps require traction (default:
stars >= 3) or explicit adoption rationale.
Reporting contract
- Keep and publish full discovery counts even when keeps are sparse.
- Emit
lane_status: pass|degraded. - Emit
degraded_causescounts using:license,stale,low-fit,archived,low-signal,no-results. - When a lane ends with
raw_count == 0after documented recovery, includeno-resultsexplicitly to keep degraded reports deterministic. - Include lane health metrics:
kept_count,raw_count,median_stars_raw,zero_star_raw. - State provenance labels used (
direct page retrieval,indexed snippet, etc.).
Rollback-safe operation
- Never delete raw evidence files.
- If stricter gates reduce keep counts too far, preserve output but mark lane degraded.
- Allow exception keeps only when rationale + risk note are explicit.
Example
- Input: 30 raw hits, many zero-star repos.
- Output: discovery evidence retained, recommendation keeps filtered by metadata/freshness/signal gates, degraded causes documented for reviewer.
Comments
Markdown supported