The docs themselves
Every page ships a plain-Markdown version alongside the rendered HTML. Use the menu in the top right of any page to copy it as Markdown, or open it directly in Claude, ChatGPT, Cursor, or VS Code. For a single file covering the whole site, point an agent at:llms.txt lists every guide and API reference page with a one-line description, so an agent can decide what to fetch instead of crawling the site.
The OpenAPI spec
For codegen, request validation, or building your own tool definitions, the machine-readable spec is:The hosted MCP server
If your AI tool speaks MCP (Claude, ChatGPT, Cursor, and most agent IDEs do), Signa runs a hosted MCP server athttps://api.signa.so/mcp. Connect it and an agent can search, retrieve, and monitor trademarks conversationally, no client code required. See MCP server to connect, and Tools & permissions for the full tool list and scopes.
The agent skill
For coding agents, install thesigna skill in the project you’re building in:
The TypeScript SDK
When an agent is writing code against Signa rather than calling it directly (a script, a backend integration, a scheduled job), point it at the@signa-so/sdk package. It’s typed end to end, so an agent gets autocomplete and compile-time checks instead of guessing field names from example JSON. See TypeScript SDK.
Which one to use
- Exploring the API or drafting a prompt: give the agent the docs or
llms.txt. - Generating a client or tool definitions: use the OpenAPI spec.
- Letting an agent act on your Signa account conversationally: connect the MCP server.
- Having a coding agent build your integration: install the agent skill.
- Writing a script or service that calls Signa: use the TypeScript SDK.