Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Fetch up-to-date library documentation using Context7 API. Use this skill when the user asks for docs, examples, or help with a specific library/framework (e.g., "look up React docs", "context7 nextjs routing", "fetch docs for fastapi").
.claude/skills/bilal140202-context7/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 39% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 40% | 0% |
This skill fetches up-to-date documentation and code examples from libraries using the Context7 API.
Context7 requires a two-step process:
First, search for the library to get its ID:
GET https://context7.com/api/v2/libs/search?libraryName=<library>&query=<context>Parameters:
libraryName: The library name (e.g., "react", "nextjs", "fastapi")query: Search context to help find the right libraryThe response contains a results array with library objects. Use the id field from the first result.
Use the library ID to fetch relevant documentation:
GET https://context7.com/api/v2/context?libraryId=<id>&query=<question>Parameters:
libraryId: The ID from step 1 (e.g., "/vercel/next.js")query: Natural language question about the documentationThis returns ranked code snippets and documentation matching your query.
When the user asks about a library:
Fetch https://context7.com/api/v2/libs/search?libraryName=nextjs&query=routing Extract the library ID from the response (e.g., "/vercel/next.js")
Fetch https://context7.com/api/v2/context?libraryId=/vercel/next.js&query=How to setup dynamic routes
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 10,555 | 13,343 | +26% | 1 | 1 | 0% | 2,159 | 2,997 | +39% | 0 | 0 | — |
case-02 | fail→fail | 10,562 | 11,300 | +7% | 1 | 1 | 0% | 2,343 | 3,030 | +29% | 0 | 0 | — |
case-03 | fail→fail | 12,660 | 15,471 | +22% | 1 | 1 | 0% | 2,677 | 3,874 | +45% | 0 | 0 | — |
case-04 | fail→pass | 5,341 | 1,575 | -71% | 1 | 1 | 0% | 1,054 | 716 | -32% | 0 | 0 | — |
case-05 | fail→pass | 5,797 | 2,539 | -56% | 1 | 1 | 0% | 1,090 | 955 | -12% | 0 | 0 | — |
case-06 | fail→pass | 7,240 | 3,100 | -57% | 1 | 1 | 0% | 1,387 | 763 | -45% | 0 | 0 | — |
case-07 | pass→pass | 7,745 | 2,084 | -73% | 1 | 1 | 0% | 1,572 | 826 | -47% | 0 | 0 | — |
case-08 | pass→pass | 11,395 | 6,902 | -39% | 1 | 1 | 0% | 2,276 | 1,936 | -15% | 0 | 0 | — |
case-09 | pass→pass | 7,402 | 2,485 | -66% | 1 | 1 | 0% | 1,332 | 877 | -34% | 0 | 0 | — |
case-10 | pass→pass | 12,777 | 6,248 | -51% | 1 | 1 | 0% | 2,184 | 1,633 | -25% | 0 | 0 | — |
case-11 | fail→pass | 3,315 | 1,619 | -51% | 1 | 1 | 0% | 530 | 740 | +40% | 0 | 0 | — |
case-12 | fail→pass | 5,851 | 1,461 | -75% | 1 | 1 | 0% | 1,007 | 724 | -28% | 0 | 0 | — |
case-17 | fail→pass | 6,905 | 3,368 | -51% | 1 | 1 | 0% | 1,414 | 1,193 | -16% | 0 | 0 | — |
case-13 | fail→pass | 9,135 | 17,137 | +88% | 1 | 1 | 0% | 1,872 | 4,012 | +114% | 0 | 0 | — |
case-14 | fail→pass | 9,526 | 13,902 | +46% | 1 | 1 | 0% | 1,968 | 3,373 | +71% | 0 | 0 | — |
case-15 | fail→pass | 10,759 | 10,066 | -6% | 1 | 1 | 0% | 2,302 | 2,548 | +11% | 0 | 0 | — |
case-16 | pass→pass | 9,135 | 6,948 | -24% | 1 | 1 | 0% | 1,913 | 1,844 | -4% | 0 | 0 | — |
case-18 | fail→pass | 9,760 | 10,111 | +4% | 1 | 1 | 0% | 1,893 | 2,617 | +38% | 0 | 0 | — |
case-19 | fail→pass | 4,355 | 6,849 | +57% | 1 | 1 | 0% | 301 | 2,056 | +583% | 0 | 0 | — |
case-20 | pass→pass | 5,555 | 10,600 | +91% | 1 | 1 | 0% | 1,196 | 2,774 | +132% | 0 | 0 | — |
case-21 | pass→pass | 6,619 | 5,720 | -14% | 1 | 1 | 0% | 1,358 | 1,584 | +17% | 0 | 0 | — |
case-22 | pass→pass | 6,341 | 5,142 | -19% | 1 | 1 | 0% | 1,368 | 1,505 | +10% | 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, and 21 counted toward the lift figure. The other 1 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 +55 percentage points is the difference between those two pass rates over the 21 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.