MCP.so
ログイン
A

Anyrun

@appcom2016

Anyrun について

概要はまだありません

基本情報

カテゴリ

その他

トランスポート

stdio

公開者

appcom2016

投稿者

appcom2016

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "anyrun": {
      "command": "python3",
      "args": [
        "-m",
        "anyrun.mcp_server"
      ]
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Anyrun?

Anyrun is a Docker sandbox execution engine for AI Agents. It provides secure, isolated code execution with full observability — automatically collecting execution traces, discovering error patterns and success paths, and extracting reusable experience as SKILL.md via LLM, with autonomous skill evolution.

How to use Anyrun?

Run any code securely with a single line of Python. All tools (sandbox_run, file operations, toolbox management) are exposed via the MCP protocol.

Key features of Anyrun

  • Zero-config Docker sandbox: one line of Python
  • Full observability: traces stored to SQLite + JSON
  • Pattern discovery: error clustering and anomaly detection
  • LLM auto-generates SKILL.md from execution patterns
  • Self-evolution: skills auto-upgrade, decay, and repair

Use cases of Anyrun

  • AI Agent tool execution
  • Secure code sandboxing
  • Agent self-improvement pipelines
  • CI/CD code execution

FAQ from Anyrun

What runtime does Anyrun require?

Anyrun requires Docker for sandbox execution and Python for client integration.

How does Anyrun store execution data?

Every execution trace is stored to SQLite and JSON for full observability.

What protocol does Anyrun use to expose tools?

Anyrun exposes all tools (sandbox_run, file ops, toolbox management) natively via the MCP protocol.

Can Anyrun evolve its skills automatically?

Yes, skills auto-upgrade, decay, and get repaired as part of the self-evolution feature.

What is SKILL.md?

SKILL.md is a file containing reusable experience automatically extracted from execution patterns by the LLM.

コメント

「その他」の他のコンテンツ