# Animica

> Animica is a live post-quantum Layer-1 blockchain (Dilithium3 + SPHINCS+
> signatures, deterministic Python smart contracts, PoIES useful-work consensus)
> with a native coin, ANM. Animica Studio is its serverless compute platform:
> run Python functions on a decentralized GPU/CPU fleet and pay in ANM.

This file helps AI agents and LLMs discover and use Animica. Everything below is
public, live, and callable.

## Animica Studio — serverless compute (Modal-style)

Write a Python function, decorate it, and run it on the fleet. Agents can use
Studio to execute and deploy compute autonomously, paying from a wallet.

- Install: `pip install animica` — the complete client. Everything to mine, run
  a node, use the wallet, deploy Python contracts, run `animica up` (the unified
  miner: PoW + useful-work + GPU train/serve + Studio functions), and use the
  Studio SDK. The native CPU miner (animica-fastpow) is included by default. This
  is what most people want.
- Install (everything): `pip install "animica[all]"` — all of the above PLUS every
  optional extra: Qt desktop-wallet QR codes, the full distributed Studio client
  (cloudpickle for closures + omni-sdk for on-chain ANM escrow), and all
  server/operator dependencies pinned. Use it for the kitchen sink or to run
  pool/API infrastructure. Quote the brackets so zsh/macOS does not glob them.
- SDK (Python):
  ```python
  import animica.studio as studio
  app = studio.App("agent-jobs")

  @app.function(image=studio.Image.debian_slim().pip_install("numpy"), gpu="A100")
  def run(seed: int) -> float:
      import numpy as np
      return float(np.random.default_rng(seed).standard_normal(1000).mean())

  run.remote(42)        # one call on the fleet (or local sandbox in dev)
  run.map(range(64))    # fan-out
  ```
- CLI: `animica studio run app.py::run --kw seed=42` · `animica studio deploy app.py`
- Modes: `ANIMICA_STUDIO_MODE=local` runs in a local sandbox with zero infra;
  `remote` escrows ANM and dispatches to the fleet.

## MCP (use Studio from an agent)

Animica ships a Model Context Protocol server exposing Studio + chain tools:

- Package: `animica-mcp` (npm) and `animica_mcp` (PyPI)
- Studio tools: `studio_run` (run a function), `studio_deploy` (register an app),
  `studio_list` (list deployed functions), `studio_estimate` (quote a run cost).
- Chain tools: balance, send, contract call, job status.
- Config example: https://github.com/animicaorg/all/tree/main/animica-mcp/config-examples

## Key links

- Studio: https://studio.animica.org
- Studio docs: https://animica.org/docs/studio
- Homepage: https://animica.org
- Explorer: https://explorer.animica.org
- JSON-RPC: https://rpc.animica.org  (namespaces: state.*, chain.*, miner.*, aicf.*)
- Source (Apache-2.0): https://github.com/animicaorg/all
- PyPI: https://pypi.org/project/animica/

## How to pay

Compute is metered per second and settled in ANM on-chain. Get ANM by mining
through the pool (https://animica.org/mine) — the only official channel today.