Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Recommend the right agents and skills for any task. Covers both heavyweight agents (Task tool) and lightweight skills (Skill tool). Triggers on: which agent, which skill, what tool should I use, help me choose, recommend agent, find the right tool.
.claude/skills/aiskillstore-tool-discovery/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | 72% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 29% | 0% |
Recommend the right agents and skills for any task.
Is this a reference/lookup task?
├── YES → Use a SKILL (lightweight, auto-injects)
└── NO → Does it require reasoning/decisions?
├── YES → Use an AGENT (heavyweight, spawns subagent)
└── MAYBE → Check catalogs belowRule: Skills = patterns/reference. Agents = decisions/expertise.
| Skill | Triggers | |-------|----------| | file-search | fd, rg, fzf, find files | | find-replace | sd, batch replace | | code-stats | tokei, difft, line counts | | data-processing | jq, yq, json, yaml | | structural-search | ast-grep, sg, ast pattern | | git-ops | git, gh, lazygit, delta, commit, PR, release, rebase | | python-env | uv, venv, pyproject | | go-ops | golang, go, goroutine, channel, context, errgroup, go test | | rust-ops | rust, cargo, ownership, tokio, serde, trait, Result, Option | | typescript-ops | typescript, type system, generics, utility types, Zod | | docker-ops | docker, Dockerfile, docker-compose, multi-stage build | | ci-cd-ops | github actions, CI, CD, pipeline, release, workflow | | api-design-ops | api design, gRPC, GraphQL, REST advanced, protobuf | | rest-ops | http methods, status codes | | sql-ops | cte, window functions | | postgres-ops | postgresql, postgres, EXPLAIN ANALYZE, vacuum, pgbouncer, JSONB, RLS, replication | | sqlite-ops | sqlite, aiosqlite | | tailwind-ops | tailwind, tw classes, dark mode, responsive | | mcp-ops | mcp server, fastmcp, tool handler, transport | | react-ops | react, hooks, useState, next.js, RSC, zustand | | vue-ops | vue, composition api, pinia, nuxt, script setup | | javascript-ops | javascript, node, esm, async/await, event loop | | astro-ops | astro, islands, content collections, partial hydration | | laravel-ops | laravel, eloquent, artisan, sanctum, pest | | payloadcms-ops | payload, payload cms, headless cms, collections | | craftcms-ops | craft, craftcms, twig, matrix fields | | asus-router-ops | asus router, asuswrt, merlin, network hardening | | nginx-ops | nginx, reverse proxy, ssl, load balancer, proxy_pass | | cloudflare-ops | cloudflare, workers, KV, D1, R2, pages, wrangler, edge | | cypress-ops | cypress, e2e, component testing, custom commands, stubbing | | bash-ops | bash, shell scripting, traps, CI scripts, defensive scripting | | claude-code-ops | claude code extensions, skills, agents, hooks, MCP, plugins | | auth-ops | jwt, oauth2, session, rbac, passkey, mfa, login | | monitoring-ops | prometheus, grafana, opentelemetry, SLO, alerting | | debug-ops | debug, crash, memory leak, race condition, bisect | | perf-ops | performance, profiling, flamegraph, bundle size, load test, benchmark | | migrate-ops | migrate, upgrade, breaking changes, codemod, version upgrade | | refactor-ops | refactor, extract, code smell, dead code, rename, restructure | | scaffold | scaffold, boilerplate, project template, init project, new project | | log-ops | JSONL, log analysis, parse logs, lnav, log search, timeline |
| Agent | Triggers | |-------|----------| | firecrawl-expert | web scraping, crawling, anti-bot | | project-organizer | restructure, organize, cleanup | | git-agent | commit, push, PR (dispatched by git-ops) | | Explore | "where is", "find" | | Plan | design, architect |
For Cloudflare/Workers, Cypress/E2E, shell scripting, Claude Code extension work, and CMS/device domains (Payload, Craft, Asus routers), use the matching -ops skill (cloudflare-ops, cypress-ops, bash-ops, claude-code-ops, payloadcms-ops, craftcms-ops, asus-router-ops). For language/framework work (Python, TypeScript, React, Postgres, etc.), use the matching -ops skill — or dispatch general-purpose with an instruction to read that skill's SKILL.md first.
Skills:
Skill tool → skill: "file-search"Agents:
Task tool → subagent_type: "firecrawl-expert"
→ prompt: "Your task"| Task | Skill First | Agent If Needed | |------|-------------|-----------------| | "How to write a CTE?" | sql-ops | — | | "Optimize this query" | postgres-ops | — | | "Find files named X" | file-search | Explore | | "Set up Python project" | python-env | — | | "What HTTP status for X?" | rest-ops | — | | "React Server Components?" | react-ops | — | | "Vue 3 composable pattern" | vue-ops | — | | "Configure nginx SSL" | nginx-ops | — | | "JWT vs session auth" | auth-ops | — | | "Set up Prometheus" | monitoring-ops | — | | "Debug memory leak" | debug-ops | — | | "Scrape a blocked site" | jina-ops | firecrawl-expert |
For complete catalogs, load:
./references/agents-catalog.md - All agents with capabilities./references/skills-catalog.md - All skills with details| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-12 | fail→pass | 6,168 | 2,346 | -62% | 1 | 1 | 0% | 1,070 | 1,837 | +72% | 0 | 0 | — |
case-01 | pass→pass | 17,635 | 2,993 | -83% | 1 | 1 | 0% | 1,935 | 1,965 | +2% | 0 | 0 | — |
case-02 | fail→pass | 12,140 | 3,322 | -73% | 1 | 1 | 0% | 1,899 | 1,902 | +0% | 0 | 0 | — |
case-03 | fail→pass | 13,938 | 4,498 | -68% | 1 | 1 | 0% | 2,188 | 2,263 | +3% | 0 | 0 | — |
case-04 | fail→pass | 10,978 | 3,291 | -70% | 1 | 1 | 0% | 2,029 | 1,991 | -2% | 0 | 0 | — |
case-05 | fail→pass | 9,217 | 2,749 | -70% | 1 | 1 | 0% | 1,478 | 1,905 | +29% | 0 | 0 | — |
case-06 | fail→pass | 13,639 | 3,286 | -76% | 1 | 1 | 0% | 2,100 | 1,975 | -6% | 0 | 0 | — |
case-07 | pass→pass | 7,708 | 2,209 | -71% | 1 | 1 | 0% | 1,209 | 1,807 | +49% | 0 | 0 | — |
case-08 | fail→pass | 8,207 | 1,810 | -78% | 1 | 1 | 0% | 1,240 | 1,738 | +40% | 0 | 0 | — |
case-09 | fail→pass | 7,277 | 3,383 | -54% | 1 | 1 | 0% | 1,157 | 1,812 | +57% | 0 | 0 | — |
case-10 | fail→pass | 12,719 | 2,370 | -81% | 1 | 1 | 0% | 2,400 | 1,779 | -26% | 0 | 0 | — |
case-11 | fail→pass | 10,026 | 2,894 | -71% | 1 | 1 | 0% | 1,517 | 1,977 | +30% | 0 | 0 | — |
case-13 | fail→pass | 16,023 | 2,459 | -85% | 1 | 1 | 0% | 2,839 | 1,851 | -35% | 0 | 0 | — |
case-14 | fail→pass | 10,212 | 2,318 | -77% | 1 | 1 | 0% | 1,599 | 1,871 | +17% | 0 | 0 | — |
case-15 | fail→pass | 11,278 | 2,829 | -75% | 1 | 1 | 0% | 2,029 | 1,933 | -5% | 0 | 0 | — |
case-16 | pass→pass | 7,495 | 2,745 | -63% | 1 | 1 | 0% | 1,262 | 1,944 | +54% | 0 | 0 | — |
case-17 | fail→pass | 7,921 | 2,132 | -73% | 1 | 1 | 0% | 1,156 | 1,871 | +62% | 0 | 0 | — |
case-18 | fail→pass | 12,774 | 1,995 | -84% | 1 | 1 | 0% | 2,353 | 1,803 | -23% | 0 | 0 | — |
case-19 | fail→pass | 8,847 | 2,036 | -77% | 1 | 1 | 0% | 1,269 | 1,801 | +42% | 0 | 0 | — |
case-20 | pass→pass | 6,767 | 4,746 | -30% | 1 | 1 | 0% | 1,283 | 2,296 | +79% | 0 | 0 | — |
case-21 | pass→pass | 8,543 | 8,088 | -5% | 1 | 1 | 0% | 1,542 | 2,913 | +89% | 0 | 0 | — |
case-22 | pass→pass | 9,173 | 6,403 | -30% | 1 | 1 | 0% | 1,565 | 2,623 | +68% | 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.