Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Tactic for vault health upkeep — lint, orphan resolution, merge candidates, confidence updates. Minimum yield ≥1 issue resolved per invocation.
.claude/skills/yogsoth-ai-vault-maintenance/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -60% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-24 | ✗→✓ | ▲ Improved | -36% | 0% |
Vault health upkeep. Called when CC judges the vault has accumulated enough entropy to warrant cleanup, or periodically between research phases.
wiki-lint-fix — run lint checks, identify issues, optionally auto-fixwiki-search — find near-duplicate candidates for mergingwiki-graph-query — explore orphan neighborhoods for reconnectionwiki-add-edge — reconnect orphans to the graph<HARD-GATE> ≥1 issue resolved per invocation. If vault_lint returns zero issues, report clean state and exit. </HARD-GATE>
CC has full autonomy over:
Typical flow (reference, not prescription):
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | wiki-add-edge | SOP wrapping vault_add_edge — create a typed relationship between two vault pages. | | wiki-edge-audit | SOP for auditing wikilink coverage — scans all edges and reports which source pages are missing dir/slug]] wikilinks to their targets. | | wiki-graph-query | SOP wrapping vault_query_graph — traverse knowledge graph from a node to explore its neighborhood and context. | | wiki-lint-fix | SOP wrapping vault_lint — run batch validation, report issues, optionally auto-fix safe problems. | | wiki-search | SOP wrapping vault_search — BM25 full-text search across vault pages. Returns ranked results with snippets. |
<!-- END available-tables (generated) -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-13 | pass→fail | 14,346 | 11,929 | -17% | 1 | 1 | 0% | 1,777 | 1,202 | -32% | 0 | 0 | — |
case-01 | fail→fail | 14,095 | 14,002 | -1% | 1 | 1 | 0% | 212 | 927 | +337% | 0 | 0 | — |
case-02 | fail→fail | 8,549 | 14,510 | +70% | 1 | 1 | 0% | 245 | 963 | +293% | 0 | 0 | — |
case-03 | fail→fail | 12,833 | 13,593 | +6% | 1 | 1 | 0% | 1,201 | 849 | -29% | 0 | 0 | — |
case-04 | fail→fail | 18,295 | 14,566 | -20% | 1 | 1 | 0% | 371 | 1,106 | +198% | 0 | 0 | — |
case-05 | fail→fail | 7,530 | 4,202 | -44% | 1 | 1 | 0% | 1,062 | 974 | -8% | 0 | 0 | — |
case-06 | pass→pass | 12,338 | 9,497 | -23% | 1 | 1 | 0% | 1,237 | 1,370 | +11% | 0 | 0 | — |
case-07 | pass→pass | 12,887 | 6,950 | -46% | 1 | 1 | 0% | 1,193 | 866 | -27% | 0 | 0 | — |
case-08 | fail→pass | 8,338 | 8,994 | +8% | 1 | 1 | 0% | 1,153 | 1,237 | +7% | 0 | 0 | — |
case-09 | pass→pass | 8,049 | 11,821 | +47% | 1 | 1 | 0% | 1,281 | 1,606 | +25% | 0 | 0 | — |
case-10 | fail→fail | 16,083 | 10,923 | -32% | 1 | 1 | 0% | 1,693 | 1,195 | -29% | 0 | 0 | — |
case-11 | fail→pass | 23,484 | 8,091 | -66% | 1 | 1 | 0% | 2,917 | 1,159 | -60% | 0 | 0 | — |
case-12 | fail→fail | 15,046 | 20,767 | +38% | 1 | 1 | 0% | 2,443 | 3,554 | +45% | 0 | 0 | — |
case-14 | fail→pass | 14,888 | 2,267 | -85% | 1 | 1 | 0% | 1,368 | 1,012 | -26% | 0 | 0 | — |
case-15 | pass→pass | 14,218 | 7,541 | -47% | 1 | 1 | 0% | 1,215 | 996 | -18% | 0 | 0 | — |
case-16 | fail→fail | 9,721 | 2,795 | -71% | 1 | 1 | 0% | 645 | 1,082 | +68% | 0 | 0 | — |
case-17 | fail→pass | 15,248 | 7,233 | -53% | 1 | 1 | 0% | 1,555 | 959 | -38% | 0 | 0 | — |
case-18 | pass→pass | 12,079 | 9,575 | -21% | 1 | 1 | 0% | 1,204 | 1,233 | +2% | 0 | 0 | — |
case-19 | pass→pass | 19,937 | 7,746 | -61% | 1 | 1 | 0% | 2,542 | 1,044 | -59% | 0 | 0 | — |
case-20 | pass→pass | 21,135 | 33,971 | +61% | 1 | 1 | 0% | 2,768 | 4,210 | +52% | 0 | 0 | — |
case-21 | pass→pass | 17,703 | 15,668 | -11% | 1 | 1 | 0% | 2,379 | 2,547 | +7% | 0 | 0 | — |
case-22 | pass→pass | 21,427 | 22,058 | +3% | 1 | 1 | 0% | 3,404 | 4,938 | +45% | 0 | 0 | — |
case-23 | pass→pass | 15,259 | 12,315 | -19% | 1 | 1 | 0% | 1,566 | 1,676 | +7% | 0 | 0 | — |
case-24 | fail→pass | 13,894 | 2,304 | -83% | 1 | 1 | 0% | 1,397 | 897 | -36% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted, and 17 counted toward the lift figure. The other 7 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +17 percentage points is the difference between those two pass rates over the 17 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.