Agent & search interfaces
LLM & agent integration
Sureshake publishes plain-text indexes, alternate Markdown representations, OpenAPI specifications, and Model Context Protocol (MCP) endpoints so answer engines and coding or research agents can read public records without scraping client UI markup.
- Root index
- /llms.txt
- Cross-origin map in llmstxt.org format
- Entity twins
- /c/{slug}.md
- Plus /p/{slug}.md & Accept: text/markdown
- Docs & API
- Markdown + OpenAPI
- docs.sureshake.com
- MCP servers
- Docs + Reports
- Public docs MCP & OAuth report MCP
Discovery & root indexes
Start at the domain root index rather than crawling navigation chromes. Both the web origin and the documentation origin publish curated Markdown indexes and JSON discovery catalogs:
- GET https://www.sureshake.com/llms.txt
- Top-level Sureshake index across product pages, public directory surfaces, documentation sections, OpenAPI specs, and MCP servers.
- GET https://www.sureshake.com/.well-known/ai-catalog.json
- Machine-readable JSON catalog (also served at /.well-known/ard.json) listing LLM indexes, OpenAPI specs, agent skills, and MCP metadata.
- GET https://docs.sureshake.com/llms.txt
- Compact documentation index linking to reader-specific section indexes and clean .md pages.
- GET https://docs.sureshake.com/llms-full.txt
- Single-document concatenation of all public docs pages (~8 MB; fetch only when full offline context is required).
Public profile Markdown twins
Every indexable public company profile (/c/[slug]) and person profile (/p/[slug]) has a companion Markdown representation assembled from the same authoritative API v2 read models as the web page.
- GET https://www.sureshake.com/c/{slug}.md
- Clean Markdown twin for a public organization or investor profile (also served at /c/{slug}/index.md).
- GET https://www.sureshake.com/p/{slug}.md
- Clean Markdown twin for a public member or Capital Graph person profile (also served at /p/{slug}/index.md).
- Accept: text/markdown
- Content negotiation on /c/{slug} and /p/{slug}: when text/markdown outranks text/html, Sureshake responds with text/markdown; charset=utf-8 and Vary: Accept.
curl -H "Accept: text/markdown, text/html;q=0.9" https://www.sureshake.com/c/sureshakeDocumentation, API & MCP
Developers and AI assistants can query public documentation, call REST API v2 endpoints directly, or connect via MCP:
- GET https://docs.sureshake.com/[...slug]/index.md
- Clean Markdown representation of any individual documentation or API reference page.
- GET https://docs.sureshake.com/reference/openapi.json
- OpenAPI 3 specification for public API v2 routes, including unauthenticated directory and public profile endpoints.
- https://docs.sureshake.com/mcp
- Public, read-only Streamable HTTP MCP server providing list_pages and get_page over product and developer documentation.
- https://api.sureshake.com/v2/mcp
- OAuth-protected read-only MCP server allowing authorized users to query and cite reports they have permission to view.
Crawling & provenance rules
Crawlers and retrieval agents must respect each origin's /robots.txt directives and per-response X-Robots-Tag headers, cache responses reasonably, and preserve attribution and provenance caveats when citing figures from Sureshake profiles or reports.
When an upstream data source is unavailable or a profile section is not covered, Sureshake endpoints state that status explicitly rather than substituting empty or synthetic figures.
Build on verified records.
Explore the public directory, read the API quickstart, or connect your assistant over MCP.