Raw: https://promptreports.ai/docs/index.md

# Docs, written for agents first

Every page has a Markdown twin an agent can fetch. A person reads the same words.

- [Quickstart](https://promptreports.ai/docs/quickstart/): Connect products, get an agent token, run your first workflow.
- [Concepts](https://promptreports.ai/docs/concepts/): Workflow, run, step, approval, record, brain, workflow request.
- [Products](https://promptreports.ai/docs/products/): What each product does in a workflow, its MCP tools and API, and a link to its own docs.
- [MCP reference](https://promptreports.ai/docs/mcp/): The PromptReports MCP server and each product's server.
- [API reference](https://promptreports.ai/docs/api/): Runs, workflows, records and usage over REST, with OpenAPI.
- [Use case guides](https://promptreports.ai/docs/guides/): Every workflow, with setup steps and its plan.
- [How agents use the docs](https://promptreports.ai/docs/agents/): Where an agent starts, what it reads, the rules it follows.
- [Machine-readable files](https://promptreports.ai/docs/machine/): llms.txt, llms-full.txt, a .md twin of every page, OpenAPI.

## Start here if you are an agent

```
GET https://promptreports.ai/llms.txt              # the company, the docs and the use cases
GET https://promptreports.ai/docs/index.md         # this page as Markdown
GET https://promptreports.ai/use-cases/index.json  # every workflow and its plan
```