Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Document ID management for Git repositories. Validates, assigns, and tracks immutable doc_ids across Markdown/YAML files. Use when working with doc_id, document IDs, "@see docid" references, document integrity validation, or shirushi commands.
.claude/skills/aiskillstore-shirushi/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -80% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -50% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -55% | 0% |
Shirushi ensures consistent, immutable document IDs across Git repositories with CI integration.
| Command | Purpose | |---------|---------| | shirushi lint | Validate doc_ids and index integrity | | shirushi scan | List documents with metadata | | shirushi show <id> | Get document info by doc_id | | shirushi assign | Assign IDs to new documents | | shirushi rehash | Recalculate content hashes |
bashshirushi lint --base HEAD~1
bashshirushi show PCE-SPEC-2025-0001-G
bashshirushi assign docs/new-spec.md
bashshirushi scan --format table
bashshirushi lint --base origin/main --check-references
Configuration is defined in .shirushi.yml:
yamlid_format: "{COMP}-{KIND}-{YEAR4}-{SER4}-{CHK1}" dimensions: COMP: type: enum values: ["API", "UI", "DB"] KIND: type: enum_from_doc_type mapping: spec: SPEC adr: ADR YEAR4: type: year digits: 4 SER4: type: serial digits: 4 scope: ["COMP", "KIND", "YEAR4"] CHK1: type: checksum algorithm: mod26AZ digits: 1
Documents use YAML frontmatter:
markdown--- doc_id: API-SPEC-2025-0001-G title: API Specification doc_type: spec --- # Content here...
Reference documents in code comments:
typescript// @see API-SPEC-2025-0001-G function handleRequest() { ... }
Shirushi tracks these references and warns when referenced documents change.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→pass | 13,327 | 19,946 | +50% | 1 | 1 | 0% | 2,481 | 2,729 | +10% | 0 | 0 | — |
case-01 | pass→pass | 11,235 | 8,154 | -27% | 1 | 1 | 0% | 1,124 | 978 | -13% | 0 | 0 | — |
case-02 | fail→pass | 25,949 | 2,675 | -90% | 1 | 1 | 0% | 4,425 | 885 | -80% | 0 | 0 | — |
case-03 | fail→pass | 17,558 | 8,394 | -52% | 1 | 1 | 0% | 2,513 | 1,045 | -58% | 0 | 0 | — |
case-04 | fail→pass | 13,796 | 2,655 | -81% | 1 | 1 | 0% | 1,595 | 905 | -43% | 0 | 0 | — |
case-05 | fail→pass | 14,574 | 6,889 | -53% | 1 | 1 | 0% | 1,624 | 812 | -50% | 0 | 0 | — |
case-06 | fail→pass | 15,699 | 6,831 | -56% | 1 | 1 | 0% | 1,791 | 808 | -55% | 0 | 0 | — |
case-07 | fail→pass | 11,615 | 7,103 | -39% | 1 | 1 | 0% | 755 | 934 | +24% | 0 | 0 | — |
case-08 | fail→pass | 6,602 | 7,393 | +12% | 1 | 1 | 0% | 1,074 | 906 | -16% | 0 | 0 | — |
case-09 | pass→pass | 15,428 | 7,361 | -52% | 1 | 1 | 0% | 1,833 | 916 | -50% | 0 | 0 | — |
case-10 | fail→pass | 12,737 | 4,181 | -67% | 1 | 1 | 0% | 2,225 | 1,315 | -41% | 0 | 0 | — |
case-11 | fail→pass | 10,703 | 7,625 | -29% | 1 | 1 | 0% | 949 | 969 | +2% | 0 | 0 | — |
case-12 | fail→pass | 15,251 | 8,054 | -47% | 1 | 1 | 0% | 1,738 | 1,000 | -42% | 0 | 0 | — |
case-13 | fail→pass | 13,729 | 2,146 | -84% | 1 | 1 | 0% | 1,385 | 878 | -37% | 0 | 0 | — |
case-14 | fail→pass | 12,514 | 4,199 | -66% | 1 | 1 | 0% | 1,341 | 1,281 | -4% | 0 | 0 | — |
case-15 | fail→pass | 12,899 | 2,191 | -83% | 1 | 1 | 0% | 1,334 | 838 | -37% | 0 | 0 | — |
case-16 | fail→pass | 15,038 | 1,957 | -87% | 1 | 1 | 0% | 1,682 | 849 | -50% | 0 | 0 | — |
case-17 | pass→pass | 19,744 | 7,044 | -64% | 1 | 1 | 0% | 2,493 | 810 | -68% | 0 | 0 | — |
case-18 | fail→pass | 13,361 | 7,234 | -46% | 1 | 1 | 0% | 1,213 | 888 | -27% | 0 | 0 | — |
case-19 | fail→pass | 17,847 | 7,017 | -61% | 1 | 1 | 0% | 1,724 | 840 | -51% | 0 | 0 | — |
case-20 | pass→pass | 13,650 | 18,164 | +33% | 1 | 1 | 0% | 2,475 | 3,060 | +24% | 0 | 0 | — |
case-22 | pass→pass | 6,153 | 11,007 | +79% | 1 | 1 | 0% | 1,100 | 1,276 | +16% | 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. 22 cases were attempted. The headline lift of +73 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.