Docs for AI agents
Read only what you need: individual Markdown pages, a compact documentation index, OpenAPI schemas, and MCP connection guides.
Docs for AI agents
Give your AI a documentation link and let it read the source. Every guide has a Markdown version with the same examples, tables, and links as the page you see here.
Connect TokPortal to your AI
Copy this into your AI conversation:
Read the TokPortal docs at https://developers.tokportal.com/mcp and guide me through connecting TokPortal to this AI app.
The MCP guide explains the connection. The integration directory has instructions for individual AI apps and coding tools.
Read a page as Markdown
Append .md to a documentation path, without the trailing slash:
| Human-readable page | Markdown source |
|---|---|
| Getting started | getting-started.md |
| Content configuration | configure-videos.md |
| MCP server | mcp.md |
| Claude integration | agents/claude.md |
| Home | index.md |
This also works for integration recipes, use cases, and blog articles. Markdown responses use Content-Type: text/markdown; each HTML page advertises its source with a rel="alternate" link. Source links inside the Markdown are absolute so they remain useful when copied into another app.
Copy page copies the Markdown, including code blocks and the canonical source URL. The adjacent Markdown link opens it directly. Neither requires an API key or a chat session.
Find the right context
Start with a small source, then expand only if the task needs it:
- Read llms.txt for the documentation map and direct Markdown links.
- Fetch the relevant guide, such as content configuration for
instagram_audio_nameor webhooks for signature verification. - Use OpenAPI when you need exact request and response schemas.
The full corpus is also available for offline indexing or broad ingestion. It is much larger than a single guide and is usually unnecessary for answering one question.
| Resource | Use it for |
|---|---|
| llms.txt | Compact map of documentation, integration guides, and machine-readable sources |
| ai-context.json | Structured API, authentication, package, operation, and documentation metadata |
| openapi.json · openapi.yaml | OpenAPI 3.1 schemas for validation and code generation |
| llms-full.txt | Full core documentation, use cases, blog posts, and OpenAPI; fetch integration recipes separately |
| developer-ecosystem.json | SDK, CLI, and MCP distribution metadata |
| package-release-manifest.json | Package release status and build metadata |
Ask the docs
Use Ask the docs for a question about TokPortal, or search by field name, endpoint, or task with ⌘K / Ctrl K. The assistant retrieves relevant sections and streams an answer with links to the documentation. Search results link directly to the matching section.
For actions in your workspace, connect your AI through MCP. The documentation assistant answers questions; it does not execute workspace operations.