# Using the API from AI agents

> API version: v1

Most callers of this API are agents, so the docs come in forms an agent can read directly. All of them are generated from the same source as these pages.

| What | URL |
|---|---|
| Index of the docs for LLMs | [`/llms.txt`](https://studyingstocks.com/llms.txt) |
| All docs in one file | [`/llms-full.txt`](https://studyingstocks.com/llms-full.txt) |
| Any docs page as Markdown | the page URL plus `.md`, for example `https://studyingstocks.com/developers/authentication.md` |
| Agent guide: which endpoint for what | [`/developers/skill.md`](https://studyingstocks.com/developers/skill.md) |
| OpenAPI 3.1 document | [`/api/v1/openapi.json`](https://studyingstocks.com/api/v1/openapi.json) |

## Tips for agents

- Give the agent the skill file and the key (as a secret, in an environment variable or the tool's credential store, never in the prompt).
- Ask for only what you need: `items=revenue,free_cash_flow&limit=5` keeps responses small.
- Branch on `error.code`, not on the message. On `rate_limited`, wait `Retry-After` seconds.
- Check `meta.generated_at` and the company's `data_updated_at` before saying how current a number is.
- Quote the `filings_url` or the SEC link when you tell a user where a number comes from.

## MCP

A Model Context Protocol server that exposes these endpoints as tools is planned. It is not available yet.