OSF APIv2 Documentation

{
  "service_type": "platform",
  "base_url": "https://osf.io",
  "auth_method": "oauth2",
  "auth_config": {
    "type": "OAuth2",
    "docs_url": "https://developer.osf.io/",
    "token_endpoint": "https://accounts.osf.io/oauth2/token",
    "authorize_endpoint": "https://accounts.osf.io/oauth2/authorize",
    "personal_access_tokens": true,
    "bearer_token_header": "Authorization: Bearer <token>"
  },
  "endpoints": [
    {
      "name": "API root",
      "url": "https://api.osf.io/v2/",
      "method": "GET",
      "description": "Root of OSF APIv2, lists top-level resources and links."
    },
    {
      "name": "Nodes",
      "url": "https://api.osf.io/v2/nodes/",
      "method": "GET",
      "description": "List and manage project nodes (projects, components)."
    },
    {
      "name": "Registrations",
      "url": "https://api.osf.io/v2/registrations/",
      "method": "GET",
      "description": "Access registrations (frozen/archived project snapshots)."
    },
    {
      "name": "Files",
      "url": "https://api.osf.io/v2/files/",
      "method": "GET",
      "description": "Waterbutler-backed file storage endpoints (osfstorage, S3, Dropbox, GitHub, etc.)."
    },
    {
      "name": "Users",
      "url": "https://api.osf.io/v2/users/",
      "method": "GET",
      "description": "User profiles and metadata."
    },
    {
      "name": "Preprints",
      "url": "https://api.osf.io/v2/preprints/",
      "method": "GET",
      "description": "Preprint submissions across OSF preprint providers."
    },
    {
      "name": "Institutions",
      "url": "https://api.osf.io/v2/institutions/",
      "method": "GET",
      "description": "Institutional affiliations and branded dashboards."
    },
    {
      "name": "Search",
      "url": "https://api.osf.io/v2/search/",
      "method": "GET",
      "description": "Search across OSF projects, registrations, preprints, files, users."
    },
    {
      "name": "Support / Documentation",
      "url": "https://osf.io/support",
      "method": "GET",
      "description": "User-facing support and help site; developer docs at https://developer.osf.io/."
    }
  ],
  "pricing_model": {
    "type": "free",
    "details": {
      "note": "OSF is a free, open-source research platform operated by the Center for Open Science (COS), a non-profit. Storage and usage are provided at no cost for standard research workflows, with institutional/enterprise partnerships available."
    }
  },
  "rate_limits": {
    "documented": false,
    "notes": "APIv2 does not publish rigid per-minute quotas in public docs. Fair-use throttling applies; large bulk fetches should paginate via JSON:API links and respect 'meta.pagination' limits (default page size 10, max 100 typically). Provide a descriptive User-Agent to avoid throttling."
  },
  "capabilities": [
    "Open-source research project management and collaboration",
    "Project and component hierarchy (nodes)",
    "Registrations and archiving of research snapshots",
    "Preprint hosting across multiple community preprint servers",
    "File storage via Waterbutler (OSF Storage, S3, Dropbox, Google Drive, GitHub, OneDrive, Box, Figshare, etc.)",
    "Add-on integrations for citations (Zotero, Mendeley), data (Dataverse), and version control (GitHub, GitLab, Bitbucket)",
    "Institutional affiliation and branded dashboards",
    "DOIs / ARKs via EZID for persistent identifiers",
    "Wikis, analytics, and contributor management",
    "OAuth2 provider (OSF acts as an OAuth2 identity provider for third-party apps)",
    "JSON:API-compliant REST API (APIv2)",
    "Search API across nodes, registrations, preprints, files, users",
    "Webhooks and add-on integrations for event-driven workflows"
  ],
  "raw_analysis": "The submitted URL (https://osf.io/support) is the front-end support/help surface of the Open Science Framework (OSF), a research collaboration and data-sharing platform operated by the non-profit Center for Open Science (COS). The HTML snippet provided is a minimal Angular single-page-app shell (osf-root, data-beasties, styles-F2YSDAXO.css, main-IJTYSFJC.js) and does not itself expose any API schema — it confirms the app is a client-rendered SPA, not a machine-readable API doc.\n\nActual API: OSF exposes a mature, publicly documented REST API — 'APIv2' — at https://api.osf.io/v2/, built on the JSON:API specification. Developer documentation lives at https://developer.osf.io/ (and historically https://developer.osf.io/documentation/v2/). The API is fully public for read operations on public resources and supports authenticated writes.\n\nAuth: OAuth2 (authorization code flow) via https://accounts.osf.io/oauth2/, plus Personal Access Tokens generated from the user's account settings. Bearer-token authentication is used on API requests.\n\nMaturity: The OSF APIv2 has been stable for years (v2 frozen since ~2015–2016), is open source (github.com/CenterForOpenScience/osf.io), and is used by thousands of researchers, i

3/3 tests passed

{
  "overall": 62,
  "dimensions": {
    "token_efficiency": 6.5,
    "first_try_success": 5.0,
    "response_parseability": 8.5,
    "error_clarity": 6.0,
    "doc_quality": 7.5,
    "auth_simplicity": 7.0,
    "latency": 9.5,
    "consistency": 7.5
  },
  "pricing_normalized": {
    "model": "free",
    "notes": "Open-source, free for standard research workflows via Center for Open Science nonprofit; institutional partnerships optional"
  },
  "issues": [
    "robots.txt explicitly blocks GPTBot, PerplexityBot, and other AI crawlers, and disallows /api/* for all user-agents — direct agent crawling/indexing is discouraged",
    "crawl-delay: 10 further throttles automated access",
    "llms.txt endpoint returns the default HTML SPA shell rather than a machine-readable AI description file",
    "Homepage <meta name=\"description\"> is empty, weakening concise value-prop summarization",
    "Authentication is standard email/password + OAuth2 rather than magic-link or agent-friendly onboarding",
    "API surface is broad (nodes, registrations, preprints, files, search, webhooks) — strong capabilities but steeper learning curve for first-try success"
  ],
  "recommendations": [
    "Publish a real /llms.txt describing OSF's API v2, auth flows, and key concepts for AI agents",
    "Add meta description / structured summary on homepage so agents can quote a crisp value prop",
    "Consider magic-link or scoped token onboarding paths for automated tools",
    "Clarify in public docs the intended 3rd-party/agent access policy given the GPTBot/PerplexityBot blocks — the API is JSON:API compliant and open, but robots.txt signals conflict with agent-friendly positioning",
    "Provide quickstart cookbook (JSON:API filters, Waterbutler file flows) to shorten time-to-first-successful-call"
  ]
}