Port of Context — Open Source Code Mode

# Port of Context (pctx) > pctx is an open-source execution layer for AI agent tool calls. > It converts MCP servers and custom tools into typed code APIs > that run in secure Deno sandboxes, replacing sequential LLM tool > calling with single-pass code execution. Self-hosted, MIT > licensed, works with any LLM. pctx is NOT an LLM, NOT an agent framework, and NOT a hosted service. It sits between your agent and your MCP servers. The core is Rust with Deno-based sandboxing. The Python SDK (pctx-py) is an HTTP client to the pctx server. - Current stable: v0.5.x, beta: v0.6.x - Install via: npm, Homebrew, or curl - SDKs: Python (pctx-py), TypeScript - Compatible with: Claude, GPT, Gemini, local models - Any existing MCP server works unchanged - Auth secrets never reach the LLM For complete documentation in a single file, see [Full Documentation](https://portofcontext.com/llms-full.txt) ## Getting Started - [README](https://raw.githubusercontent.com/portofcontext/pctx/main/README.md): Installation, quick start, architecture overview, usage modes (HTTP server and unified MCP server) - [Code Mode](https://raw.githubusercontent.com/portofcontext/pctx/main/docs/code-mode.md): How Code Mode works — presenting MCP servers as code APIs instead of direct tool calls ## CLI & Configuration - [CLI Reference](https://raw.githubusercontent.com/portofcontext/pctx/main/docs/CLI.md): All pctx commands — init, start, mcp add/dev/start - [Configuration](https://raw.githubusercontent.com/portofcontext/pctx/main/docs/config.md): Config file format, auth, allowed hosts, ToolDisclosure settings - [Upstream MCP Servers](https://raw.githubusercontent.com/portofcontext/pctx/main/docs/upstream-mcp-servers.md): Adding and managing upstream MCP server connections ## Python SDK - [Python SDK](https://raw.githubusercontent.com/portofcontext/pctx/main/pctx-py/README.md): @tool decorator, AsyncTool classes, Pydantic schemas, agent framework integration ## Sandbox & Runtime - [Code Exec