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

prompt-caching.ai
March 13, 2026
1 min read
46/100
Summary
Open source MCP plugin that automatically injects prompt cache breakpoints into Claude Code sessions. Up to 90% token cost reduction — zero config.
Coverage & discussions
Hacker News · prompt-caching.ai
Community discussions
Community Sentiment
Positives
- The auto-caching feature introduced by Anthropic significantly simplifies the process for developers, allowing for efficient breakpoint placement without manual intervention.
- Reducing token usage by 90% through prompt-caching can lead to substantial cost savings, making AI applications more accessible for developers.
- Claude Code's built-in prompt caching enhances user experience by automatically managing system prompts and conversation history, streamlining development.
Concerns
- The plugin discussed is outdated and does not provide any incremental benefits over the existing auto-caching feature from Anthropic, which diminishes its relevance.
- Despite initial perceptions, Sonnet is not cheaper than Opus when considering output token costs, highlighting potential marketing discrepancies that can mislead users.
Related Articles

I burned all my tokens researching how to save tokens
Jul 19, 2026

Tmux and Claude Code: Sessions, Send-Keys, and the Ceiling
Oct 19, 2026

Superlite: A Faster Alternatie to Superpowers
Oct 19, 2026

How the Claude Code team does code review
Oct 16, 2026

CEO fired developers to make room for AI. Developers create open source AI CEO
Aug 27, 2026