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

# Connect Claude

> Add Audity as a remote MCP connector inside Claude Desktop or claude.ai.

## Requirements

* Claude account on Pro, Max, Team, or Enterprise (Custom Connectors require a paid plan)
* An Audity Personal Access Token (`aky_...`), see [Authentication](/authentication)

## Claude Desktop / claude.ai

<Steps>
  <Step title="Open Connectors">
    Settings → **Connectors** → **Add Custom Connector**.
  </Step>

  <Step title="Enter the URL">
    ```
    https://docs.auditynow.com/mcp
    ```
  </Step>

  <Step title="Configure auth">
    The connector needs `Authorization: Bearer aky_<your-token>` on every request to the Audity backend.

    * If Claude shows an "Authorization header" or "API key" field on the connector setup screen: paste `Bearer aky_<your-token>` (the full string including `Bearer `, since Claude does not auto-prefix in this flow).
    * If Claude redirects you through an OAuth approval first: approve the redirect. After approval, you'll still get a prompt for your Audity API key. Paste your `aky_<your-token>` (this time without `Bearer `, the connector wraps it).

    The exact UX depends on Claude Desktop's current connector flow, which Anthropic iterates on. If you see something unexpected, reach for the [troubleshooting section](#troubleshooting).
  </Step>

  <Step title="Test">
    Start a new chat. Ask:

    > List my Audity projects.

    Claude should call `listProjects` and return your real projects. If it asks for permission first, approve.
  </Step>
</Steps>

## Claude Code (CLI)

Claude Code has its own dedicated guide that covers both the MCP config file approach and the direct REST API fallback. See [Claude Code guide](/guides/claude-code).

## What you get

The connector exposes Audity's full agent surface as Claude tools, **28 operations** across projects, leads, Nucleus, and account. Tool names match the `operationId` in the [OpenAPI spec](/api-reference/openapi.json):

* `getCurrentUser`, `getCurrentTier`, `getCredits`
* `listProjects`, `createProject`, `getProject`, `patchProject`
* `triggerAuditAnalysis`, `getAuditAnalysis`
* `listOpportunities`, `getDeliverables`
* `listLeads`, `getLead`, `convertLead`
* `listMemories`, `createMemory`, `deleteMemory`
* `listCaptures`, `createCaptureNote`, `getCapture`, `reprocessCapture`, `deleteCapture`
* `listContacts`, `createContact`, `updateContact`, `deleteContact`
* `listInsights`, `getSuggestions`

Plus the docs themselves are searchable as context, Claude can answer "how do I X with Audity" by reading the docs page directly without calling an API.

## Common prompts

```text theme={null}
List my Audity projects.
```

```text theme={null}
Search my Audity Nucleus memories for {topic}.
```

```text theme={null}
Start a new audit for {client}, {industry}, {size}. Run the analysis. Tell me when it's done.
```

```text theme={null}
Convert my top 3 Audity leads from this week into projects, after confirming with me.
```

```text theme={null}
Pull the deliverables for project {id}. Format the executive summary as a client-ready brief.
```

## Tips for agent prompts

* **Tell Claude to confirm before spending credits.** Audit analysis costs credits, lead conversion costs 1,000 each. Without explicit confirmation in the prompt, Claude will sometimes fire off a batch you didn't intend.
* **Ask Claude to cite IDs.** When summarizing opportunities or memories, have Claude include the `id` field in its output. That way you can drill into a specific item and verify Claude isn't confabulating.
* **Use the `confidence` field on memories.** Tell Claude to flag any memory under 0.8 confidence as a hypothesis rather than a fact. Detected patterns are AI-generated; treat them accordingly.

## Troubleshooting

<AccordionGroup>
  <Accordion title="Claude says it can't find Audity tools">
    Check the connector status in Settings → Connectors. If it's red, re-authorize. If your PAT was revoked or expired, generate a new one and update the connector.
  </Accordion>

  <Accordion title="Connector hangs on OAuth approval">
    Open `https://docs.auditynow.com/authed/mcp/oauth/token` in a browser to confirm the docs connector is reachable. If the page errors, try again in 5 minutes as there may be a brief service hiccup.
  </Accordion>

  <Accordion title="Every call returns 401 PAT_MALFORMED">
    Token format is wrong. Most common cause: the connector double-prefixed `Bearer Bearer aky_...`. Edit the connector auth and remove one `Bearer `, or paste only `aky_...` (depends on which UX field you're filling, see Step 3 above).
  </Accordion>

  <Accordion title="Every call returns 403 PAT_SCOPE_INSUFFICIENT">
    Token lacks `write` scope. Generate a new token with both `read` and `write` scopes, then update the connector.
  </Accordion>

  <Accordion title="403 errors on every call but token is fine">
    Your plan may not include the operation you're trying to run. Some plans are read-only or lead-generation-focused, while audit-capable paid plans unlock writes. Check `GET /api/user/tier` and any `requiredTier` field in the 403 response.
  </Accordion>

  <Accordion title="429 rate limited mid-batch">
    See [Authentication → Rate limits](/authentication#rate-limits). Write limits and long-running audit synthesis are the most common ceilings for batch workflows.
  </Accordion>

  <Accordion title="Audit analysis times out">
    `triggerAuditAnalysis` runs **synchronously** for 60–300 seconds. If Claude reports a timeout, the call may have succeeded server-side. Use the async audit tool only when the project already has current document and interview analyses, or verify with `getAuditAnalysis` before re-triggering. Re-triggering an in-flight analysis wastes credits.
  </Accordion>
</AccordionGroup>

## What's next

* [Run a full audit workflow →](/guides/running-an-audit)
* [Lead conversion playbook →](/guides/lead-conversion)
* [Working with Nucleus →](/guides/working-with-nucleus)