Code Execution

Shared, consistent permission-prompt and policy layer for AI coding agent tools

Model Context Protocol (MCP) server for Sandock - Sandbox in Docker for AI Agents

观远 BI 虚拟工作站 MCP Server - WASI 沙箱内提供 shell 工具链与观远 BI CLI，对外暴露 MCP Streamable HTTP 端点

iota embedded policy engine as a pi coding-agent extension: hash-pinned bless, agents.yaml enforcement for read/write/edit/bash, JSONL decision log

The flintd daemon and CLI: one local server that gives every agent harness on the machine the same library of tools.

Minimal Slack agent harness built on the pi AI libraries.

UV-centric Jupyter kernel execution for Claude via MCP with enhanced error handling

Zero-trust security gateway for MCP (Model Context Protocol) tool servers — verifies a tool's declared capabilities against observed kernel syscalls, then confines it per call with mounts, seccomp and network namespaces.

No description declared by any source.

Interactive REPL for testing any MCP server over stdio or HTTP

No description declared by any source.

MCP server for Jupyter notebooks

Runs Python in a local, hardware-isolated micro-VM that keeps its state between calls

A cross-platform sandboxed file system operations toolkit based on Model Context Protocol (MCP)

Named Docker and Fly Machine sandbox computers with an interactive MCP App dashboard.

MCP server providing persistent SageMath sessions (Sage REPL over the Model Context Protocol)

No description declared by any source.

mcp-run-python-code: MCP server for running Python code, installing packages, and executing Python files.

A secure and sandboxed python code interpreter for LLMs. Accessable via MCP, or directly in code

MCP server providing a restricted Python execution tool