CloudLLM-ai

Memory that lasts and compounds. MentisDB gives agents durable memory so they do not just remember, they improve over time. It stores append-only thought chains plus a Git-like skills registry, letting skills evolve with experience through versioned, integrity-checked updates and fast retrieval.

17
2
100% credibility
Found Mar 23, 2026 at 17 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

MentisDB is a persistent memory system and skill library for AI agents that stores thoughts in a tamper-proof chain and versions skills like code updates.

How It Works

1
📰 Discover a smarter way for AI

You hear about MentisDB, a helper that gives your AI assistants a lasting memory so they never forget important thoughts or lessons.

2
📥 Get the memory helper

Download and set up the simple program on your computer with one easy command.

3
▶️ Wake up the shared brain

Start the program and it quietly runs in the background, ready to store everything your AI learns.

4
🔗 Connect your AI buddy

Tell your favorite AI coding tool, like Claude or Copilot, to use this new memory by adding a quick link.

5
💭 Watch memories build

As you chat with your AI, it saves key ideas, decisions, and skills automatically into the permanent brain.

6
📊 Peek at the dashboard

Open a web page to browse all saved thoughts, agent details, and skill updates in a friendly view.

🧠 AI remembers forever

Your AI picks up right where it left off, shares knowledge across tools and team members, and gets smarter over time.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 17 to 17 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is mentisdb?

MentisDB is a Rust-built durable memory engine for AI agents, storing append-only thought chains and a Git-like versioned skills registry so knowledge compounds across sessions. It runs as a daemon (mentisdbd) with HTTP MCP, REST APIs, and a web dashboard, letting agents like those in Claude Code or GitHub Copilot CLI share a persistent brain that survives context resets and tool swaps. Think memory that lasts forever for LLMs—agents remember decisions, lessons, and skills without ephemeral chat history.

Why is it gaining traction?

Unlike volatile LLM contexts or basic github memory allocators, MentisDB offers hash-chained permanence with fast queries, multi-agent coordination, and signed skill updates, integrating seamlessly via MCP for memory github copilot or memory github llm workflows. Developers hook it up in seconds (cargo install mentisdb; claude mcp add), getting fleet-scale orchestration where agents bootstrap from recent_context and evolve skills without losing ground. Benchmarks show solid throughput even at 10k concurrent tasks.

Who should use this?

AI agent builders running multi-tool setups like Cursor, VS Code agents, or Qwen Code, especially those hitting github memory limits in long-running tasks. Orchestrators of agent fleets needing shared append-only memory for project handoffs, retrospectives, or self-improving skills. LLM devs tired of re-explaining context every session.

Verdict

Promising for agent memory that outlives chats, but at 17 stars and 1.0% credibility, it's early—docs are solid with MCP/REST examples, full tests/benches, yet low adoption signals caution. Try for prototypes if persistent agent brains fit; watch for maturity.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.