6/ Another underrated feature: You can actually inspect what your agent remembers. Every interaction is visible through the dashboard. When an agent behaves unexpectedly, being able to trace its…

pip install memanto: @moorcheh_ai #AgenticMemory #AIInfrastructure

Most devs try to fix this the wrong way. They stuff CLAUDE. md with instructions. They paste context into every prompt. They build brittle memory features inside their agent. The result? Fragile.…

Everything is logged and inspectable via dashboard..... super useful when debugging weird agent behavior. Open source (Apache 2.0 / MIT), public GitHub, PRs welcome. Built to be useful, not to lock…

I built a small Memanto bridge for Claude Code skills so context can survive across /grill-with-docs, /tdd, /diagnose, and /handoff. It recalls project memories before a skill starts, then stores…

The interesting engineering: • Multi-anchor semantic search to surface a full namespace (Memanto's recall caps at top-100 by similarity) • Line-based KIND::CONTENT extraction (way more reliable than…

Built a LangGraph + Memanto example for durable cross-session memory. Session 1 stores support memories. Session 2 starts with a fresh thread and still recalls the order + replacement preference. PR:…

Small Memanto example for Claude Code / mattpocock-style skills: recall project decisions before a skill run, pass them through MEMANTO_SKILL_CONTEXT, then save new durable notes from the run. PR:…

30-second demo for my LangGraph + Memanto bounty submission. It shows: - Day 1 memories stored in Memanto - A fresh LangGraph run recalling those memories - Long-term memory kept outside LangGraph…

Submitted LangGraph + Memanto cross-session memory agent for $100 bounty @moorcheh_ai 2 3 Agent remembers facts across conversations. No API key needed. 4 5 6 7 Star the repo to help! 8 9 #Memanto…