Docs for AI agents
These docs are also published in machine-readable formats for LLMs and coding agents. Point your agent at any of these resources.
| Resource | URL | Use it for |
|---|---|---|
llms.txt | /developers/api/llms.txt | An index of every page with a one-line summary. |
llms-full.txt | /developers/api/llms-full.txt | Every page’s full text in one file, ready to drop into context. |
| Page Markdown | Add .md to any page URL, like /developers/api/quickstart.md | One page as clean Markdown. |
| OpenAPI 3.1 | /developers/api/openapi/bloomtext-api.yaml | Generating clients, tools, and typed SDKs. Includes the OAuth scheme and scopes. |
| Agent skill | /developers/api/skills/bloomtext-api/SKILL.md | Teaching Claude Code, Codex, and other agents the API’s rules. See Agent skill. |
Copy any page
Use the Copy page menu at the top of any page to copy it as Markdown, or to open it in ChatGPT or Claude and ask questions about it.
Give your agent the whole API
Use the BloomText API docs at https://www.bloomtext.com/developers/api/llms-full.txt and the OpenAPI spec at https://www.bloomtext.com/developers/api/openapi/bloomtext-api.yaml. For patient operations, my server uses an organization API key from BLOOMTEXT_API_KEY as a Bearer token; every patient operation names its patient by ID, phone number, or MRN. For an org admin's own conversations, my app uses OAuth (authorization code with PKCE), with its client ID and secret in BLOOMTEXT_CLIENT_ID and BLOOMTEXT_CLIENT_SECRET. Send an Idempotency-Key on every POST. Help me build the integration.
Agent skill
The BloomText API skill teaches a coding agent the API’s rules: OAuth and scopes, search syntax, idempotency, pagination, error codes, and what never to send. Agents that support Agent Skills load it automatically for tasks that involve BloomText.
Claude Code
mkdir -p .claude/skills/bloomtext-api
curl -o .claude/skills/bloomtext-api/SKILL.md https://www.bloomtext.com/developers/api/skills/bloomtext-api/SKILL.mdTo have an AI agent read and act in BloomText for your users, instead of writing code against the API, see Build an AI agent.