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

# MCP server

> One connection for expertise, managed tasks, evidence, and saved results.

CrowdListen exposes three tools:

| Tool | Purpose |
| - | - |
| `recall` | Read authorized workspaces, expertise, task status, and evidence |
| `analyze` | Quote, run, or cancel reviewed expertise; analyze evidence |
| `ingest` | Save knowledge, sources, observations, and coverage |

[Set up your client](/docs/setup/agents) before making requests. The installed tool schemas are the complete contract for supported modes and fields.

## Remote MCP

The remote endpoint is `https://mcp.crowdlisten.com/mcp`. It supports authenticated HTTP clients and exposes the same three tools. The client must securely supply a valid CrowdListen account session using an `Authorization: Bearer` header and manage its refresh.

The endpoint currently does not implement OAuth or anonymous URL-only setup. A successful `initialize` or `tools/list` response does not establish authorized data access. Verify `recall({ mode: "connection" })` from the connected client. Never paste session credentials into chat. Most local clients should use the browser-authenticated stdio harness in the [setup guide](/docs/setup/agents).

## Expertise

Create and review a private method in [My expertise](https://crowdlisten.com/expertise). Choose an authorized workspace, then read the method and its revision:

```js theme={null}
recall({ mode: "workspaces" })
recall({ mode: "expertise", workspace_id: "WORKSPACE_UUID" })
recall({ mode: "expertise", expertise_id: "EXPERTISE_UUID" })
```

Quote the input before running it. Quotes do not run a model or debit credits.

```js theme={null}
analyze({
  action: "expertise",
  operation: "quote",
  expertise_id: "EXPERTISE_UUID",
  expected_revision: 1,
  files: [{ name: "conversation.txt", content: "Selected, authorized text" }]
})
```

Check `ready`, `maximumCredits`, `availableCredits`, `expiresAt`, and `quoteHash`. After the user authorizes that scope and maximum charge, submit the same revision and files:

```js theme={null}
analyze({
  action: "expertise",
  operation: "run",
  expertise_id: "EXPERTISE_UUID",
  expected_revision: 1,
  files: [{ name: "conversation.txt", content: "Selected, authorized text" }],
  request_id: "NEW_TASK_UUID",
  quote_hash: "RETURNED_QUOTE_HASH",
  confirmed: true
})
```

Preserve the request ID if the response is uncertain. Read progress rather than starting another task:

```js theme={null}
recall({ mode: "expertise_task", task_id: "TASK_UUID" })
```

The managed task continues after the client disconnects. Retrieve the published `artifactId` and exact file revision from its manifest:

```js theme={null}
recall({ mode: "expertise_result", package_id: "ARTIFACT_UUID" })
recall({
  mode: "expertise_result",
  package_id: "ARTIFACT_UUID",
  revision: 1,
  path: "README.md"
})
```

Result reads are free. Use the returned revision and a listed path; do not guess filenames. These reads do not launch a new analysis or change sharing.

Cancel with:

```js theme={null}
analyze({ action: "expertise", operation: "cancel", task_id: "TASK_UUID" })
```

Work already performed can consume credits. Uncertain usage remains reserved until reconciled. Workspace access and method ownership are checked on each call.

## Research and collection

Inspect subjects with `recall({ mode: "entities" })`. Verify a chosen subject using `recall({ mode: "connection", entity_id: "ENTITY_UUID" })`. This checks current scoped reads and permissions; it does not prove live platform access.

`recall({ mode: "collection_plan", entity_id: "ENTITY_UUID" })` reads the saved source scope. An agent with authorized browser or source tools can collect originals, save them with `ingest({ destination: "sources", ... })`, and record coverage separately. See the [collection skill](https://crowdlisten.com/skills/collect-sources/SKILL.md) for the complete schema and retry rules.

## Pilot experts

Some first-party experts use an additional protected router connection. Read the [expert skill](https://crowdlisten.com/skills/experts/SKILL.md), reuse the user's CrowdListen login, and check account-specific availability. Discovery and installation must not start paid work.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.