ccproxy CLAUDE.md

Project instructions for CCProxy, a Go program that translates requests between Claude Code and different AI providers.

In plain words
What is it for?
Use them when working on CCProxy's command-line interface, server, request pipeline, provider integrations, routing, or translation code.
Why use it?
They explain the project structure and security details needed to change the proxy without misunderstanding its role or requirements.

Instructions file

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/orchestre-dev/ccproxy/claude-md
Clone the repo
git clone --depth 1 https://github.com/orchestre-dev/ccproxy
Per session 3,327 This file is loaded in full into every session.
When invoked 3,327 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.03327 $0.03327
Opus 5 $0.01664 $0.01664
Sonnet 5 $0.00665 $0.00665
Haiku 4.5 $0.00333 $0.00333

Measured 2d ago against content hash 36c99273e5ec, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ccproxy CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 381 lines

How it starts

The opening of the file, as written. The whole thing — 381 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CCProxy - AI Request Proxy for Claude Code (Go Implementation)

Project Overview

CCProxy is a high-performance Go-based API translation proxy that enables Claude Code to work with multiple AI providers. It acts as a middleware layer that translates Anthropic's API format to various provider-specific formats, enabling seamless integration without modifying Claude Code. This is a complete rewrite from TypeScript to Go with enhanced security, performance, and reliability features.

Important: CCProxy is a translation proxy - it does not add new capabilities beyond API format translation. It requires models with function calling support for Claude Code compatibility.

Architecture

Core Components

  • CLI Interface (cmd/ccproxy/) - Cobra-based command-line interface
  • Server (internal/server/) - Gin-based HTTP server with middleware stack
  • Pipeline (internal/pipeline/) - Request processing pipeline with transformations
  • Providers (internal/providers/) - Multi-provider AI service integration
  • Transformers (internal/transformer/) - Request/response transformation engine
  • Router (internal/router/) - Intelligent model routing based on token count and parameters
  • Performance (internal/performance/) - Monitoring, rate limiting, and resource management
  • Security (internal/security/) - Authentication, authorization, and audit logging

Key Features

  • API Translation: Converts Anthropic API format to provider-specific formats
  • Intelligent Routing: Automatic model selection based on:
    • Token count (>60K → longContext route)
    • Model type (haiku models → background route)
    • Boolean thinking parameter (thinking: true → think route)
    • Explicit provider selection (format: "provider,model")
  • Provider Support: Full transformers for Anthropic, OpenAI, Gemini, DeepSeek, OpenRouter
  • Streaming Support: Server-Sent Events (SSE) for real-time responses
  • Process Management: Background service with PID file locking and graceful shutdown
  • Claude Code Integration: Auto-start, environment variable management, reference counting

Read the full file on GitHub · 381 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 381 lines · 3,327 tokens per session scan A 36c99273e5ec

Subscribe to this mod's changes

ccproxy CLAUDE.md is an instructions file published in the GitHub repository orchestre-dev/ccproxy (24 stars, last pushed 1y ago), licensed MIT. It adds 3,327 tokens to every session, about $0.0166 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.