elidickinson/pi-claude-bridge

Claude Code inference provider for Pi. Allows the use of Claude Code of Pro/Max subscriptions from pi.dev.

What it solves

This project provides a bridge between the pi coding agent and Claude Code, allowing users to use Claude models (Opus, Sonnet, Haiku) as the primary intelligence for pi while maintaining the pi TUI (Terminal User Interface) and tool-calling capabilities. It also allows pi to delegate specific tasks to Claude Code as a separate tool when using other providers.

How it works

The bridge operates in two primary modes:

  1. Provider Mode: It registers Claude models as providers within pi. When a user sends a request, the bridge forwards the request and tool calls to Claude Code via the Agent SDK, ensuring that bash commands and skills are handled correctly. It manages session synchronization between pi's history and Claude Code's sessions.
  2. AskClaude Tool: This is an optional tool that allows pi's current LLM (regardless of the provider) to delegate tasks or questions to Claude Code. This can be configured for read-only access or "full" mode (read, write, and bash access).

Who it’s for

Developers using the pi coding agent who want to integrate Claude's latest models and the specific capabilities of the Claude Code CLI tool into their workflow.

Highlights

  • Seamless Model Integration: Automatically pulls Claude model lists from pi-ai's catalog, supporting high-context windows (up to 1M tokens) for supported plans.
  • Flexible Delegation: The AskClaude tool allows for second opinions, code reviews, or autonomous task handling via Claude Code.
  • TUI Preservation: All tool calls and tool results flow through pi's TUI, maintaining a consistent user experience.
  • Detailed Debugging: Includes comprehensive logging for both the bridge and the Claude Code CLI subprocesses to troubleshoot session-resume failures.
  • Cofigurable Control: Users can customize the plan type, context window limits, and the path to the Claude Code executable.

Related

  • Project
    wesammustafa/Claude-Code-Everything-You-Need-to-Know

    A practical, up‑to‑date guide for Anthropic’s Claude Code CLI, covering installation, prompt engineering, custom slash‑commands (Skills), automatic Hooks, sub‑agents, JavaScript workflows, and the Model‑Context‑Protocol for external tool integration. Includes ready‑to‑copy `.claude/` examples for teams building AI‑augmented development pipelines.

  • Project
    UNLINEARITY/CLI-WeChat-Bridge

    CLI WeChat Bridge is a Node‑JS tool that links local AI‑coding CLIs (Codex, Claude Code, OpenCode, Pi) to personal or Enterprise WeChat. It forwards chat messages to the chosen CLI, captures output, and sends replies back, preserving the native terminal experience. Features include daemon mode, file transfer, rich slash commands, emoji bindings, and cross‑platform support, all under an AGPL‑3.0‑or‑later license.

  • Dispatch
    Claude Code Cross-Session Messaging: How It Works, When to Use It, and Security Considerations

    Claude Code’s cross-session messaging lets one Claude session send plain‑text alerts or answers to another, enabling coordinated parallel work while respecting per‑session permission controls.

  • Dispatch
    Claude Code Best Practices Guide

    Anthropic provides a comprehensive set of best practices for optimizing Claude Code, an agentic coding environment, focusing on context window management and autonomous verification.

  • Project
    RichardAtCT/claude-code-openai-wrapper

    A production‑ready server that makes Claude Code models look like the OpenAI API, supporting chat, streaming, sessions, tool use, multi‑provider auth, and Docker deployment.