{
  "data": null,
  "kind": "anchor.page",
  "links": {
    "api": "https://www.anchorterminal.com/api/v1/index.json",
    "html": "https://www.anchorterminal.com/niches/code.library-docs/",
    "json": "https://www.anchorterminal.com/niches/code.library-docs/index.json",
    "llms": "https://www.anchorterminal.com/llms.txt",
    "markdown": "https://www.anchorterminal.com/niches/code.library-docs/index.md",
    "mcp": "https://www.anchorterminal.com/mcp",
    "slim": "https://www.anchorterminal.com/niches/code.library-docs/index.min.md"
  },
  "markdown": "No company's docs show a full fit yet. The listings below do the job with a limit, or not at all.\n\n| Companies, full fit | With a limit | Carry the tag | Share of the category's work | First-hand accounts |\n| --- | --- | --- | --- | --- |\n| 0 | 5 | 11 | 22 | 18 |\n\n## What counts as done\n\nA doc excerpt for the right library and version, by resolve then query, showing the API as it exists in the version in use, with a not-found result that is distinguishable from an error.\n\nMeasured, for our own tests: the excerpt is under 4,000 tokens and comes in at most two calls.\n\nHow agents ask for it: use context7; check the docs for this library's latest version; how do I call X in version 6.1; find a code sample for this Azure API\n\n## Does the job, with a limit\n\nIt does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.\n\n- **amCharts 5 MCP Server (amCharts)**: get_doc, get_api_reference and search_docs return amCharts 5 pages and class references with no key, but only for amCharts 5, with no version parameter, no size cap and no documented error shape.\n- **Context7 (Upstash)**: resolve-library-id then query-docs, version-suffixed ids such as /vercel/next.js/v16.2.9, 404 version_not_found and isError on failures; but no size parameter, ESP-IDF 6.1 missing three months after release (#3753), and a free user was told to create a key mid-session.\n- **GitMCP (Ido Salomon (GitMCP open-source project))**: fetch and search documentation tools per public repository with no key; but public repositories only (#248), no version parameter, a tool-name mismatch in search results (#258), and no commit or maintainer reply since 8 May 2026.\n- **Microsoft Learn MCP Server (Microsoft Corporation)**: microsoft_docs_search, microsoft_docs_fetch and microsoft_code_sample_search with language, no key, and ?maxTokenBudget caps search; but fetch returns whole pages that hit the model's limit (#109), search served a stale chunk (#228), and there is no version parameter.\n- **Ref (Ref Software, Inc)**: ref_search_documentation then ref_read_url is two calls, pages are trimmed to relevant sections and ref_src=private reaches a team's repositories and PDFs; but a key needs a browser sign-up, the free 200 credits never renew, and there is no version or size parameter.\n\n## Couldn't check\n\nThe docs were closed to us, or silent. These carry no mark either way.\n\n- **ModusToolbox AI Assistant (Infineon Technologies AG)**: The listing describes a library catalogue, board discovery and code examples for Infineon devices, but no tool names, schemas or limits are published and the implementation source is private, so the two-call, 4,000-token criterion cannot be read.\n\n- not this job: Azure DevOps MCP Server (Microsoft Corporation): No documentation tools; Microsoft Learn MCP is the docs server.\n- not this job: Entire (Entire Inc): No documentation tools.\n- not this job: Git (MCP reference server) (Model Context Protocol, a Series of LF Projects, LLC): No documentation tools.\n- not this job: GitHub MCP Server (GitHub, Inc.): github_support_docs_search answers GitHub product questions only; no library documentation.\n- not this job: GitLab MCP Server (GitLab Inc.): No documentation tools.\n- not this job: graft (Vishisht Choudhary (individual)): No documentation tools.\n- not this job: Graphify (Graphify Labs): Indexes the repository's own docs and PDFs, not library documentation by version.\n- not this job: Jam MCP (Jam): No documentation tools.\n- not this job: Momentic MCP server (Momentic): momentic_search_docs searches Momentic's own docs only.\n- not this job: REA (morluto): No documentation tools.\n- not this job: Salesforce DX MCP Server (Salesforce, Inc.): LWC guidance tools return fixed documentation prompts; not a versioned library docs lookup.\n- not this job: Sourcegraph MCP Server (Sourcegraph, Inc.): Searches the company's indexed code, not library documentation.\n- not this job: Superpowers (Prime Radiant (no legal form or registration found on its pages); the MIT LICENSE names Jesse Vincent): No documentation tools.\n- not this job: TokenEconomizer (Microsoft Corporation (repository in the microsoft GitHub organisation with Microsoft's SECURITY.md); co-authors at UC Berkeley): No documentation tools.\n- not this job: Void (Cloudflare, Inc. (VoidZero)): Ships its own docs in node_modules/void/docs for agents; not a library docs tool.\n\n## What stops agents finishing\n\n- index lags the release (ESP-IDF 6.1 missing three months after release)\n- anonymous rate limit hit mid-session and the model asked to create a key\n- fetch of a long page exceeds the model's token budget with no size parameter\n- the model's own safety layer flags fetched docs as injection\n- public repositories only, no private docs (GitMCP)\n\n## Around it\n\nBefore: agent-harnesses.pair-in-terminal. After: code.commit-from-sandbox.\n\n## Where this comes from\n\nWeight 22 of 100 in Code \u0026 developer platforms: 18 of 65 accounts (Context7 10, GitMCP 3, DeepWiki 3, Microsoft Learn 2); the largest single cluster found and the one users most often call out as the MCP server they keep\n\nAs data: https://www.anchorterminal.com/api/v1/use-cases/code.library-docs.json, or the MCP tool get_use_case with id code.library-docs. Draft, read 2026-11-12 from public docs and first-hand accounts, not yet reviewed by a person.\n",
  "meta": {
    "attribution": "Anchor Terminal (https://www.anchorterminal.com)",
    "attributionRequired": true,
    "credit": "CC BY 4.0, credit Anchor Terminal and link the page or API address you took it from (https://www.anchorterminal.com/terms/#data)",
    "docs": "https://www.anchorterminal.com/docs/",
    "generatedAt": "2026-11-16",
    "license": "CC-BY-4.0",
    "method": "https://www.anchorterminal.com/benchmark/",
    "methodology": "0.8",
    "openapi": "https://www.anchorterminal.com/openapi.json",
    "preview": false,
    "run": "2026-10-01",
    "runLabel": "October 2026 research run",
    "terms": "https://www.anchorterminal.com/terms/#data"
  },
  "page": {
    "breadcrumbs": [
      {
        "name": "Home",
        "url": "https://www.anchorterminal.com/"
      },
      {
        "name": "For agents",
        "url": "https://www.anchorterminal.com/agents/"
      },
      {
        "name": "Niches",
        "url": "https://www.anchorterminal.com/niches/"
      },
      {
        "name": "Fetch current library documentation before writing code",
        "url": ""
      }
    ],
    "description": "Fetch current library documentation before writing code: 0 companies fully do it, 5 with a limit, 11 carry the tag. Every listing in Code \u0026 developer platforms with the reason and a source.",
    "facts": null,
    "h1": "Fetch current library documentation before writing code",
    "image": "https://www.anchorterminal.com/assets/og/default.png",
    "path": "/niches/code.library-docs/",
    "published": "",
    "section": "agents",
    "title": "Fetch current library documentation before writing code | Anchor…",
    "toc": null,
    "updated": "2026-11-15",
    "url": "https://www.anchorterminal.com/niches/code.library-docs/"
  },
  "tokens": {
    "markdown": 1500,
    "slim": 280
  },
  "version": 1
}
