> ## Documentation Index
> Fetch the complete documentation index at: https://docs.talentreview.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent API overview

> How AI agents discover, read, and act on TalentReview.ai jobs

TalentReview\.ai exposes a small REST surface designed for AI agents: browse and
search jobs, read a single job, apply on behalf of a user, and post new roles.
It is the same platform recruiters use, reachable over plain HTTP with JSON.

## Base URL

```text theme={null}
https://api.talentreview.ai
```

## Discovery

An agent can find everything it needs without being told the endpoints up front:

<Columns cols={2}>
  <Card title="llms.txt" icon="robot" href="https://www.talentreview.ai/llms.txt">
    A curated index of the machine-readable surfaces: job Markdown twins, the
    skill doc, and the MCP server.
  </Card>

  <Card title="Skill doc" icon="book" href="https://api.talentreview.ai/agents/skill.md">
    The `/agents/*` endpoints, auth, and request shapes in Markdown.
  </Card>

  <Card title="Skills manifest" icon="list" href="https://www.talentreview.ai/.well-known/skills/index.json">
    Declares the <code>talentreview-jobs</code> skill.
  </Card>

  <Card title="Job Markdown" icon="file-lines">
    Append <code>/.md</code> to any public job URL, e.g.
    <code>[www.talentreview.ai/jobs/](http://www.talentreview.ai/jobs/){'{'}id{'}'}/.md</code>, for a clean text version.
  </Card>
</Columns>

## MCP server

For agents that speak the Model Context Protocol, the `trai-recruiter` MCP
server exposes recruiting tools natively (`list_jobs`, `get_job`,
`list_applications`, `get_candidate`, `get_kanban_board`,
`update_application_status`, `score_application`, `get_dashboard`,
`analyze_candidate`, `generate_interview_questions`) over SSE, authorized with
OAuth 2.0 via Auth0.

* SSE endpoint: `https://mcp.talentreviewai.com/sse`
* Manifest: [`/.well-known/mcp.json`](https://www.talentreview.ai/.well-known/mcp.json)
* OAuth discovery: `https://mcp.talentreviewai.com/.well-known/oauth-authorization-server`

## What to read next

<Columns cols={2}>
  <Card title="Authentication" icon="key" href="/agents/authentication">
    When you need an API key, and how to send it.
  </Card>

  <Card title="Endpoints" icon="code" href="/agents/endpoints">
    Browse, search, read, apply, and create jobs.
  </Card>

  <Card title="Errors" icon="triangle-exclamation" href="/agents/errors">
    The structured error shape and status codes.
  </Card>

  <Card title="SDK & CLI" icon="terminal" href="/agents/sdk">
    The <code>talentreview</code> npm package: typed client and CLI.
  </Card>
</Columns>