Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Programmatic discovery of agent SKILL.md files under configured public repo roots (infrastructure, scripts, projects/templates, resource-pool templates, docs/prompts, .agents/skills, and .cursor/skills). Use when enumerating skills, validating .cursor/skill_manifest.json, writing docs/_generated/skills_index.md, checking docs/prompts workflow contracts, or wiring editor/Codex/Hermes/MCP automation. Exposes discover_skills, write_skill_manifest, manifest_matches_discovery, and check_skill_contrac
.claude/skills/docxology-infrastructure-skills/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -72% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 27% | 0% |
Machine-readable discovery for all SKILL.md descriptors used by Cursor and similar agents.
**/SKILL.md under the default roots infrastructure/, scripts/, projects/templates/, fonds/templates/, rules/templates/, tools/templates/, docs/prompts/, .agents/skills/, and .cursor/skills/ (configurable).agents/skills/<name>/SKILL.md files are included so Codex/Hermes descriptors also appear in .cursor/skill_manifest.json and MCP list_skillsname and description (full dict available).cursor/skill_manifest.jsondocs/_generated/skills_index.md for human browsingdocs/prompts/**/SKILL.md metadata (version, last_updated, status, data_access_level, task_type, modes, related_skills).agents/skillscollection into Codex, Claude Code, and Hermes with python -m infrastructure.skills runtime-status|runtime-install
bashuv run python -m infrastructure.skills write uv run python -m infrastructure.skills write-index uv run python -m infrastructure.skills write --roots infrastructure docs/prompts .cursor/skills uv run python -m infrastructure.skills check uv run python -m infrastructure.skills check-contracts uv run python -m infrastructure.skills list-json uv run python -m infrastructure.skills runtime-status
discovery.py — API implementationcontracts.py — docs/prompts workflow metadata contractsinfrastructure/SKILL.md — human-oriented hub list| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 3,809 | 4,210 | +11% | 1 | 1 | 0% | 506 | 664 | +31% | 0 | 0 | — |
case-02 | fail→fail | 13,124 | 3,959 | -70% | 1 | 1 | 0% | 2,022 | 699 | -65% | 0 | 0 | — |
case-03 | fail→fail | 4,952 | 3,533 | -29% | 1 | 1 | 0% | 742 | 632 | -15% | 0 | 0 | — |
case-04 | fail→pass | 17,749 | 2,270 | -87% | 1 | 1 | 0% | 2,236 | 913 | -59% | 0 | 0 | — |
case-05 | fail→pass | 18,159 | 2,574 | -86% | 1 | 1 | 0% | 3,174 | 890 | -72% | 0 | 0 | — |
case-06 | fail→pass | 6,051 | 2,517 | -58% | 1 | 1 | 0% | 1,149 | 914 | -20% | 0 | 0 | — |
case-07 | fail→pass | 10,025 | 2,855 | -72% | 1 | 1 | 0% | 1,862 | 1,019 | -45% | 0 | 0 | — |
case-08 | fail→fail | 10,270 | 5,819 | -43% | 1 | 1 | 0% | 1,933 | 712 | -63% | 0 | 0 | — |
case-09 | fail→pass | 3,922 | 1,713 | -56% | 1 | 1 | 0% | 582 | 737 | +27% | 0 | 0 | — |
case-10 | fail→pass | 6,410 | 2,058 | -68% | 1 | 1 | 0% | 1,153 | 772 | -33% | 0 | 0 | — |
case-11 | fail→pass | 7,422 | 1,593 | -79% | 1 | 1 | 0% | 1,108 | 743 | -33% | 0 | 0 | — |
case-12 | pass→pass | 6,717 | 1,987 | -70% | 1 | 1 | 0% | 1,117 | 720 | -36% | 0 | 0 | — |
case-13 | fail→pass | 6,518 | 3,395 | -48% | 1 | 1 | 0% | 1,115 | 823 | -26% | 0 | 0 | — |
case-14 | fail→pass | 6,672 | 2,181 | -67% | 1 | 1 | 0% | 1,178 | 837 | -29% | 0 | 0 | — |
case-15 | fail→pass | 6,777 | 3,537 | -48% | 1 | 1 | 0% | 1,224 | 1,100 | -10% | 0 | 0 | — |
case-16 | fail→pass | 8,310 | 2,639 | -68% | 1 | 1 | 0% | 1,428 | 968 | -32% | 0 | 0 | — |
case-17 | pass→pass | 7,912 | 3,179 | -60% | 1 | 1 | 0% | 1,385 | 1,033 | -25% | 0 | 0 | — |
case-18 | fail→pass | 7,410 | 2,092 | -72% | 1 | 1 | 0% | 1,250 | 787 | -37% | 0 | 0 | — |
case-19 | fail→pass | 9,140 | 2,369 | -74% | 1 | 1 | 0% | 1,639 | 876 | -47% | 0 | 0 | — |
case-20 | fail→pass | 7,428 | 2,157 | -71% | 1 | 1 | 0% | 1,301 | 872 | -33% | 0 | 0 | — |
case-21 | fail→pass | 10,643 | 2,041 | -81% | 1 | 1 | 0% | 1,454 | 786 | -46% | 0 | 0 | — |
case-22 | pass→pass | 11,522 | 9,040 | -22% | 1 | 1 | 0% | 2,003 | 2,349 | +17% | 0 | 0 | — |
case-23 | pass→pass | 16,747 | 6,892 | -59% | 1 | 1 | 0% | 2,300 | 1,801 | -22% | 0 | 0 | — |
case-24 | fail→pass | 9,631 | 6,887 | -28% | 1 | 1 | 0% | 1,663 | 1,758 | +6% | 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 20 counted toward the lift figure. The other 4 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 +67 percentage points is the difference between those two pass rates over the 20 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.