Reference · verified 2026-08-14

Animica MCP server — blockchain tools for AI agents

Q: What are the honest limitations?

The AI inference tool (animica_ai_ask) rides on community-GPU capacity, so responses can be slow or unavailable when no worker is serving the requested model — check animica_ai_models first. Chain tools are read-only by design: an agent that needs to send transactions must use the full animica package with its own wallet instead.

pip install animica-mcp gives any Model Context Protocol client — Claude Code, Claude Desktop, Cursor, VS Code — 15 read/compute tools for the Animica post-quantum blockchain: chain head/block/account lookups, OpenAI-compatible AI inference, a verifiable quantum randomness beacon, mining-pool stats, and Studio compute estimates. It is read-only by design: no private keys, no signing, no spending, zero required configuration.

Install

pip install animica-mcp     # console command: animica-mcp
# or run without installing:
uvx animica-mcp

animica-mcp is a thin console wrapper around animica[mcp]; the server implementation lives in the animica package (animica.mcp.server, also runnable as animica mcp serve).

Client setup

Claude Code

claude mcp add animica -- uvx animica-mcp

Claude Desktop (`claude_desktop_config.json`)

{
  "mcpServers": {
    "animica": { "command": "uvx", "args": ["animica-mcp"] }
  }
}

Cursor (`~/.cursor/mcp.json` or project `.cursor/mcp.json`)

{
  "mcpServers": {
    "animica": { "command": "uvx", "args": ["animica-mcp"] }
  }
}

VS Code (`.vscode/mcp.json`)

{
  "servers": {
    "animica": { "type": "stdio", "command": "uvx", "args": ["animica-mcp"] }
  }
}

If you installed with pip, replace "command": "uvx", "args": ["animica-mcp"] with "command": "animica-mcp".

The 15 tools

Tool	What it does
`animica_info`	Orientation: what Animica is, endpoints, docs. Start here.
`animica_ai_ask`	Ask Animica's AI a question (OpenAI-compatible chat inference).
`animica_ai_models`	List available AI models.
`animica_ai_job_status`	Status of an AI request/job by id (read-only).
`animica_quantum_beacon_latest`	Latest round of the verifiable quantum randomness beacon.
`animica_quantum_draw`	Verifiable quantum-random draw off a beacon round (lottery, weighted choice, shuffle, range, coin, dice, bytes).
`animica_quantum_verify`	Re-verify a quantum draw client-side — pure offline compute.
`animica_qdna_verify_gene`	Verify a qDNA training-genome gene seal — pure offline compute.
`animica_chain_head`	Chain head: height, hash, chain id. Read-only.
`animica_chain_block`	Read a block by height or hash. Read-only.
`animica_chain_account`	Balance + nonce for an `anim1…` address. No keys, no signing.
`animica_pool_stats`	Live mining-pool stats.
`animica_network_hashrate`	Current network hashrate.
`animica_studio_estimate`	Estimate the ANM cost of an Animica Studio run — local quote, charges nothing.
`animica_studio_functions`	List functions deployed to Animica Studio. Read-only.

Configuration (all optional)

Every backend defaults to public Animica infrastructure — the server works with zero configuration. Environment variables if you want to override:

Variable	Default	Purpose
`ANIMICA_BASE_URL`	`https://pool.animica.org/v1`	OpenAI-compatible inference base. `https://animica.dev/v1` is the free keyless endpoint if you prefer it.
`ANIMICA_RPC_URL`	`https://rpc.animica.org/rpc`	Node JSON-RPC (read-only calls).
`ANIMICA_POOL_URL`	`https://pool.animica.org`	Mining-pool stats API.
`ANIMICA_BEACON_URL`	`https://pool.animica.org`	Quantum randomness beacon.
`ANIMICA_API_KEY`	(empty)	Optional Bearer key for inference; not required.
`ANIMICA_MCP_TIMEOUT_S`	`60`	HTTP timeout for backend calls.
`MCP_TRANSPORT`	`stdio`	`stdio`, `streamable-http` (serves 127.0.0.1:8765), or `sse`.

Security posture & honest limits

Read + compute only. Chain access goes through a hard-coded allow-list of read-only JSON-RPC methods; mutating methods (send, sign, deploy, claim) are not on the list and cannot be reached, even by a bug.
No keys. The server never handles private keys, never signs, never spends, and never mutates wallet state. Agents that need to transact must use the full animica package with their own wallet.
Inference depends on community capacity. animica_ai_ask rides on community GPU workers; when no worker serves the requested model, responses can be slow or unavailable. Call animica_ai_models first.
Quantum draw/verify and qDNA verification are pure local compute and work fully offline.

FAQ

What is the Animica MCP server?

A Model Context Protocol server for the Animica post-quantum blockchain and AI network. Installing animica-mcp (pip install animica-mcp, or run ad hoc with uvx animica-mcp) gives any MCP client 15 read/compute tools: chain head, block, and account lookups, OpenAI-compatible AI inference, a verifiable quantum randomness beacon, mining-pool stats, and Studio compute cost estimates.

Can the Animica MCP server spend funds or sign transactions?

No. It is read + compute only. Chain access goes through a hard-coded allow-list of read-only JSON-RPC methods, so mutating methods (send, sign, deploy, claim) cannot be reached even by a bug. The server never handles private keys, never signs, and never spends.

Which MCP clients does it work with?

Any MCP client. Tested configs are documented for Claude Code (claude mcp add animica -- uvx animica-mcp), Claude Desktop, Cursor, and VS Code. The default transport is stdio; streamable-http and legacy SSE are available via the MCP_TRANSPORT environment variable.

Do I need an API key or a wallet to use it?

No. All backends default to public Animica infrastructure and the server works with zero configuration. An optional ANIMICA_API_KEY environment variable exists for the inference API but is not required, and no wallet is involved.

What are the honest limitations?

The AI inference tool (animica_ai_ask) rides on community-GPU capacity, so responses can be slow or unavailable when no worker is serving the requested model — check animica_ai_models first. Chain tools are read-only by design: an agent that needs to send transactions must use the full animica package with its own wallet instead.

Canonical docs

PyPI: animica-mcp · PyPI: animica
gitlab.com/Animica/animica-core (Apache-2.0)
Agents hub — chain, inference, and payments quickstarts
Free AI inference API — the endpoint behind animica_ai_ask
Post-quantum blockchain API — the RPC behind the chain tools