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

# CreatorAudit for AI agents

> Drive CreatorAudit from an LLM agent over MCP or the v2 API — typed, bearer-authenticated, with a machine-readable OpenAPI contract and curated tools.

CreatorAudit's API is built to be driven by software, including LLM agents. It gives an agent everything it needs to act reliably without hand-holding: a single bearer key, a documented OpenAPI contract, predictable JSON envelopes, RFC 9457 errors, cursor pagination, and a request ID on every response.

If you are wiring CreatorAudit into an agent, autonomous workflow, or MCP-based assistant, this section is for you.

## Why it's agent-ready

* **One credential.** Authenticate every call with `Authorization: Bearer YOUR_API_KEY`. The key is scoped to one organization, so an agent never has to reason about per-resource access. Create keys on the [API keys](https://app.creatoraudit.com/app/api-keys) page and confirm with `GET /v2/whoami`.
* **A machine-readable contract.** The full OpenAPI 3.1 document is downloadable from the [API reference](/api-reference/introduction) and served at `https://api.creatoraudit.com/v2/openapi.json`. Turn it into tool or function definitions automatically — no scraping, no guessing.
* **Predictable envelopes.** Every collection returns `{ data: [...], pagination: { next_cursor, has_next, limit } }`, so one parser handles every list endpoint.
* **Structured errors.** Failures are [RFC 9457](https://www.rfc-editor.org/rfc/rfc9457) problem details (`application/problem+json`) with a machine-readable `code` field — easy to branch on in a tool loop.
* **Traceability.** Every response carries an `X-Request-ID` header. Log it so a human (or another agent) can trace any call.

<Note>
  Build against `v2` (`https://api.creatoraudit.com/v2`). `v1` is the legacy, deprecated
  surface and exists only for older integrations.
</Note>

## What an agent can do

With a single key, an agent can:

* **Track accounts** — list, add, update, and remove the Instagram and TikTok accounts your organization follows (`/accounts`).
* **Track videos** — track individual posts by URL and browse videos discovered on tracked accounts (`/videos`, `/account-videos`).
* **Group creators** — organize accounts into creators and manage their membership (`/creators`).
* **Read analytics** — pull daily metrics and time series for accounts, videos, and creators (`/analytics/...`, `/{resource}/{id}/analytics`).
* **Get leaderboards** — rank your top creators over a period (`/creators/top`).
* **Read the headline view** — fetch organization-level figures (`/overview`).
* **Search** — find accounts, creators, and videos by query (`/search`).

The [API reference](/api-reference/introduction) covers the shared conventions, and the [interactive explorer](https://api.creatoraudit.com/v2/docs) lists every endpoint with its live schema.

## Explore the agent guides

<Columns cols={2}>
  <Card title="Use the API as agent tools" icon="wrench" href="/agents/tool-use">
    Turn the OpenAPI spec into tool/function definitions for OpenAI- and Anthropic-style tool calling, and execute calls safely.
  </Card>

  <Card title="Connect via MCP" icon="plug" href="/agents/mcp">
    Run the official CreatorAudit MCP server to drive accounts, videos, creators, and
    analytics from your agent — or search these docs over MCP.
  </Card>

  <Card title="Agent recipes" icon="list-check" href="/agents/recipes">
    End-to-end workflows — onboard and report, daily standup, find and track, backfill
    metrics.
  </Card>

  <Card title="Docs as agent context" icon="book" href="/agents/context">
    Feed these docs to your agent via <code>llms.txt</code>, per-page Markdown, the docs MCP, and the OpenAPI spec.
  </Card>
</Columns>

## These docs are agent-friendly, too

This documentation site is designed to be read by machines as well as people:

* **`/llms.txt`** and **`/llms-full.txt`** (also under `/.well-known/`) give an agent a curated map or a full-text dump of the docs.
* **Append `.md` to any page URL** to get a clean Markdown version of that page.
* The **official CreatorAudit MCP server** wraps the full v2 API as agent tools (every endpoint), so your agent can manage accounts, videos, creators, and analytics with just your API key. It's hosted — point your client at the URL with your key, or install the one-click Claude Desktop bundle. See [Connect via MCP](/agents/mcp).
* A **docs MCP server** at the site's `/mcp` path exposes search and content-retrieval tools, so your agent can query these docs live.

See [Docs as agent context](/agents/context) for the exact URLs and when to use each.

<Tip>
  Questions about building on CreatorAudit? Reach out at
  [support@creatoraudit.com](mailto:support@creatoraudit.com).
</Tip>