mcpField
A
anima

anima

Remote
useanima.shCommunication2Updated 5w ago
RepositoryVisit website

About anima

anima is a hosted MCP server that gives AI agents email, phone, SMS, voice, and credential-management capabilities. It exposes the Anima platform through one gateway and domain-specific endpoints.

Key features

  1. 01Manage agent identities and lifecycle
  2. 02Create and manage email inboxes
  3. 03Send mail and handle threads
  4. 04Provision numbers and manage SMS
  5. 05Handle voice calls, transcripts, and recordings
  6. 06Store credentials and manage OAuth connections

Use cases

  1. 01Set up a dedicated email address for an AI agent
  2. 02Handle communication tasks through SMS and voice
  3. 03Store and retrieve authentication credentials

MCP support

The hosted MCP server exposes its tools to clients over streamable HTTP with Bearer authorization. The /mcp endpoint provides the full tool set, alongside domain-specific endpoints.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://mcp.useanima.sh/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

@anima-labs/mcp-server

The hosted MCP server behind https://mcp.useanima.sh — the recommended way to connect AI assistants to Anima. It exposes the full tool surface (66 tools today) over streamable HTTP with Authorization: Bearer <ak_… or mk_…>.

This package is deliberately not published to npm: it is the deployed gateway. For a local/stdio server use @anima-labs/mcp (published, 53 core tools) — same platform, same auth. Client setup for both lives in the MCP docs, or run anima setup-mcp.

Endpoints

One gateway, scoped mounts. /mcp carries everything; the per-domain mounts serve tailored connections with a smaller tool count:

  • /mcp — all tools (recommended)
  • /agent — agent lifecycle (create/get/list/update/delete)
  • /email — Anima agent mailboxes: inbox create/get/list/update/delete, send/reply/forward, threads, drafts, attachments, custom domains (agents.useanima.sh addresses or your own domain — no Gmail/SMTP relay involved)
  • /phone — phone number provisioning, SMS threads, voice calls + transcripts/recordings (Telnyx-backed)
  • /platform — account/usage overviews and webhook management
  • /vault — encrypted credential vault: CRUD, search, TOTP, server-side use, credential requests, OAuth connections, Connect Links
  • /extension — browser extension connect (headless / Puppeteer)

Run locally (development)

bun run src/index.ts

Environment: ANIMA_API_URL (defaults to a local dev API at http://127.0.0.1:3100; the production deployment sets it to the live API), MCP_BASE_URL, PORT.

Deploy production (mcp.useanima.sh)

Preferred path is GitHub Actions: run Deploy MCP Server (.github/workflows/deploy.yml) via workflow_dispatch.

Required repository secrets (for Workload Identity Federation / OIDC):

  • GCP_WORKLOAD_IDENTITY_PROVIDER (format: projects/<number>/locations/global/workloadIdentityPools/<pool>/providers/<provider>)
  • GCP_SERVICE_ACCOUNT_EMAIL (deployer service account in anima-labs)

Workflow behavior:

  • Builds and pushes us-central1-docker.pkg.dev/anima-labs/anima/mcp-server:<tag>
  • Deploys Cloud Run service mcp-server in us-central1
  • Prints GET /health response so you can confirm startedAt changed

Manual fallback (same image/deploy settings) remains:

./deploy.sh <tag>

History

Replaces five separate packages (mcp-agent, mcp-email, mcp-phone, mcp-platform, mcp-vault) and the mcp-core shared library.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.