원클릭으로
detecting-staleness
Scan all artifacts for confidence metadata and report staleness using pure derived scoring.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Scan all artifacts for confidence metadata and report staleness using pure derived scoring.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Capture point-in-time MER (MiDi Experience Record) for a wallet — data state, visual screenshot, perception context, and decision metadata.
Compare user expectations (UTCs) with code reality to identify gaps. Use when you need to understand discrepancies between what users expect and what code actually does.
Batch scan canvases for GAP sections, deduplicate, route to correct repos, draft issue bodies, and file confirmed issues.
Parallel multi-user processing using Claude Code native teams. Leader spawns workers per user, each runs /ingest-dm, leader aggregates for cross-canvas patterns.
Validate gap hypotheses by manually simulating features for individual users and measuring commitment, not satisfaction.
Automated end-to-end feedback pipeline that pulls new Supabase entries, enriches, classifies, routes to canvases, and generates a synthesis report.
| name | detecting-staleness |
| description | Scan all artifacts for confidence metadata and report staleness using pure derived scoring. |
| user-invocable | true |
| allowed-tools | Read, Write, Glob, Grep, Bash |
Scan all artifacts with confidence metadata, compute freshness scores using a pure derived formula, and output a staleness report with suggested actions.
Confidence is always derived, never stored. The score is computed from immutable inputs (frontmatter + git history + config). Same inputs always produce the same score.
/stale # All artifacts, default threshold
/stale --threshold 0.50 # Custom threshold
/stale --pack observer # Specific pack's artifacts
Glob the following paths for artifacts with confidence frontmatter:
grimoires/*/canvas/*.md
grimoires/*/journeys/*.md
grimoires/*/reality/*.md
grimoires/artisan/taste.md
IF --pack specified:
Filter to grimoires/{pack}/ only
For each file:
Parse YAML frontmatter
Look for confidence: block
When frontmatter is MISSING: Classify artifact as unknown with suggested action:
Run `/refresh {path}` to initialize confidence tracking
For each artifact WITH confidence metadata:
READ INPUTS (from frontmatter):
last_validated_commit # Git SHA anchor
validation_count # Times re-validated
READ INPUTS (from git):
commits_since = git rev-list --count {last_validated_commit}..HEAD -- {related_paths}
IF last_validated_commit not found in history:
Fallback: git log --oneline --since="30 days ago" -- {related_paths} | wc -l
Warn: "Baseline commit not found — using 30-day window"
IF .git unavailable (shallow clone, no git):
Use validation_count only
Mark score as "approx"
Warn: "Git history unavailable — confidence approximate"
READ INPUTS (from .loa.config.yaml):
decay_rate = confidence.decay.rate (default: 0.02)
code_change_penalty = confidence.penalties.code_change (default: 0.05)
revalidation_boost = confidence.boosts.revalidation (default: 0.20)
diminishing_factor = confidence.boosts.diminishing_factor (default: 0.7)
floor = confidence.decay.floor (default: 0.20)
initial_score = confidence.initial_score (default: 0.90)
COMPUTE:
days_since = days between last_validated_commit date and now (UTC)
time_decay = days_since * decay_rate
code_penalty = commits_since * code_change_penalty
boost = revalidation_boost * (diminishing_factor ^ validation_count)
score = clamp(initial_score - time_decay - code_penalty + boost, floor, 0.95)
CLASSIFY:
score >= healthy_threshold (0.70) → "healthy"
score >= warning_threshold (0.50) → "warning"
score >= stale_threshold (0.30) → "stale"
score < stale_threshold → "archive"
STALENESS REPORT
════════════════
| Artifact | Confidence | Status | Last Validated | Suggested Action |
|----------|-----------|--------|----------------|------------------|
| canvas/xabbu.md | 0.72 | healthy | 2026-02-01 | — |
| canvas/elcapitan.md | 0.48 | warning | 2026-01-20 | /drift, /refresh |
| reality/profile.md | 0.31 | stale | 2026-01-10 | /ground profile |
| taste.md | unknown | unknown | — | /refresh grimoires/artisan/taste.md |
Summary: {N} healthy, {M} warning, {P} stale, {Q} archive, {R} unknown
IF --threshold specified: Only show artifacts below the threshold.
Thresholds and decay parameters are read from .loa.config.yaml:
confidence:
initial_score: 0.90
decay:
rate: 0.02
floor: 0.20
thresholds:
healthy: 0.70
warning: 0.50
stale: 0.30
penalties:
code_change: 0.05
boosts:
revalidation: 0.20
diminishing_factor: 0.7
If config is missing, use the defaults shown above.
| Error | Resolution |
|---|---|
| No artifacts found | Report "No artifacts with confidence metadata found" |
| Git not available | Use validation_count only, mark as "approx" |
| Config missing | Use defaults |
| Frontmatter parse error | Skip artifact, log warning |
| Baseline commit not in history | Use 30-day window fallback |
.loa.config.yaml/drift (detailed per-artifact), /refresh (re-validate)