🚀 Autocache - Intelligent Anthropic API Cache Proxy Automatically inject cache-control fields into Claude API requests to reduce costs by up to 90% and latency by up to 85%. Works as a transparent drop-in replacement for popular AI platforms like n8n, Flowise, Make.com, LangChain, and LlamaIndex—no code changes required

Local reverse proxy between Claude Code CLI and AWS Bedrock — fixes prompt caching TTL, Agent SDK missing cache, thinking/effort optimization, and model ID mapping

**Pruner** is a lightweight CLI wrapper for Claude Code that automatically reduces your API token usage through local proxy interception, context pruning, and prompt caching — without changing how you work.

Proxy that lets Claude Code talk to self-hosted sglang/vLLM backends with KV cache normalization and vision routing

Claude Code Cache Bug Fix using mitmproxy!

Claude Max OAuth proxy — unlocks 1M context, prompt caching, and API-only features for Claude Code on Replit and beyond

Normalizing HTTP proxy between Claude Code and vLLM — rips out cache busters for stable KV cache prefix (22% → 95%+ hit rate)

自动给 anthropic claude 模型添加缓存，可结合 one api 等项目将 anthropic claude API 转换成自带缓存的 OpenAI 格式 API -- 此项目主要由 claude code + glm-4.6 进行开发和维护

claude-cache-proxy module from Projects collection

an auto-cache proxy for claude

Local Squid caching proxy for Claude Code sessions

Project based course to build a Cache Proxy, built with claude code teams, for learning purposes

OpenAI-compatible proxy for Claude Code CLI — use your Claude Pro/Max subscription as an API endpoint. Supports sessions, prompt caching, streaming (SSE), and extended thinking.

Serverless proxy backend for SHealth iOS app — sits between the app and Anthropic Claude API for key security, caching, and rate limiting

Cost-optimizing reverse proxy for Claude/LLM APIs. Routes each request to the cheapest model that can handle it. Semantic caching, DAG workflows, health-aware backends, ML-assisted routing.

An AI-first development environment for solo developers. Run multiple AI coding agents (Claude, Gemini) in parallel across tmux sessions, with all the infrastructure you need to ship — database, cache, object storage, email, reverse proxy, and monitoring.