Dev benchmarks OptMem vs Claude Code native memory for 5 weeks; open-source Rust agent memory engine surfaced
doodlestein · x · 2026-09-15
LangWatch dev @rchaves shared a 5-week analysis of replacing Claude Code's native memory with Victor Taelin's OptMem. Using LangWatch's full session tracking, he ran a natural A/B comparison (some sessions randomly didn't activate OptMem), finding OptMem works better for his use cases.
In the thread, @doodlestein promoted his open-source eideticenginecli: a durable, local-first, explainable memory substrate for coding agents, written in Rust on his "franken-stack" (Asupersync, FrankenSQLite, Frankensearch, FrankenNetworkX), with tests, fuzzing, proofs, and skills directories (44 stars).
Key points:
- Real-world agent memory comparison backed by weeks of tracked usage
- A ready-to-try open-source memory engine CLI for coding agents
Related event: Five-Week Test: OptMem vs Claude Code's Native Memory(3 posts)→
More from coding & agent
- Man runs agent to auto-post on his FB page every 30 min overnight, using just 1% of quota — TheMoonMidas · 2026-09-15
- AgentBridge: open-source compatibility layer to swap agent frameworks painlessly — 0sparsh2 · 2026-09-15
- Ant Group's HazardAuditor adds execution-grounded safety supervision for computer-use agents — antgroup · 2026-09-15
- AistyMCP: open-source per-tool permissions for MCP servers, deny-by-default — iamjoehoward · 2026-09-15
- Local Qwen loops and forgets in coding agents while Claude Code just works — tlpta · 2026-09-15
- Pareta routes cheap LLM tasks to small models, 620x cheaper than GPT-5.5 — D33B · 2026-09-15