DigiNews

Tech Watch by Johan Denoyer

← Back to articles

Show HN: Agent-cache – Multi-tier LLM/tool/session caching for Valkey and Redis

Quality: 8/10 Relevance: 8/10

Summary

Show HN: Agent-cache presents a multi-tier cache for AI agents backed by Valkey and Redis, caching LLM responses, tool results, and session state behind a single connection. It includes framework adapters (LangChain, LangGraph, Vercel AI SDK) and built-in observability (OpenTelemetry, Prometheus), with v0.1.0 shipped and v0.2.0 adding cluster mode; streaming support is planned. The project targets breaking framework-locks by providing a unified caching layer across agents and sessions.

🚀 Service construit par Johan Denoyer