Prompt-caching – auto-injects Anthropic cache breakpoints (90% token savings)

Source: Hacker News

MCP Plugin · Open Source · MIT
Built with Claude, for Claude.

How prompt‑caching works

Anthropic’s caching API stores stable content server‑side for 5 minutes. Cache reads cost 0.1× instead of 1×. This plugin places the breakpoints automatically.

🐛 BugFix Mode

Detects stack traces in your messages. Caches the buggy file + error context once. Every follow‑up only pays for the new question.

♻️ Refactor Mode

Detects refactor keywords + file lists. Caches the before‑pattern, style guides, and type definitions. Only per‑file instructions are re‑sent.

📂 File Tracking

Tracks read counts per file. On the second read, injects a cache breakpoint. All future reads cost 0.1× instead of 1×. (always on — all modes)

🧊 Conversation Freeze

After N turns, freezes all messages before turn (N − 3) as a cached prefix. Only the last 3 turns are sent fresh. Savings compound.

Benchmarks

Measured on real Claude Code sessions with Sonnet. Break‑even at turn 2.

Cache creation costs 1.25× normal. Cache reads cost 0.1×. Every turn after the first is pure savings.

Install prompt‑caching

Claude Code (recommended)

⏳ Pending approval in the official Claude Code plugin marketplace. Install directly from GitHub in the meantime:

/plugin marketplace add https://github.com/flightlesstux/prompt-caching
/plugin install prompt-caching@ercan-ermis

Claude Code’s plugin system handles everything automatically. The get_cache_stats tool is available immediately after install.

Install globally via npm

npm install -g prompt-caching-mcp

Add to your client’s MCP config

{
  "mcpServers": {
    "prompt-caching-mcp": {
      "command": "prompt-caching-mcp"
    }
  }
}

Supported MCP‑compatible clients include Cursor, Windsurf, ChatGPT, Perplexity, Zed, Continue.dev, and any other MCP client.

Open source · MIT · Zero lock‑in

Ready to cut your Claude Code token costs by 90%?