Pmbai
Stores decisions, lessons, and project facts in a local SQLite file, automatically recalls relevant memory before agent reasoning, and journals agent work after, with hybrid recall and a visual graph/timeline dashboard.
Launch screenshots from Product Hunt, Jun 2026
A local-first memory layer for AI coding agents, solving the problem of agents forgetting context between sessions by storing decisions, lessons, and project facts in a single SQLite file on the user's disk. It targets developers and engineering teams using AI coding tools like Claude Code, Cursor, Codex, and Zed, and is positioned as an open-source, offline solution with no API keys or cloud dependency, offering fast recall via MCP integration.
Key features
- Auto-recall on every prompt
- Sub-millisecond async writes
- Hybrid recall with BM25 + dense vectors
- Entity graph visualization
- Timeline journal with git-graph
- Lessons scored by follow-rate
- MCP integration for multiple agents
- Offline, no API keys, no cloud
Social posts
- No social media activity within the last 30 days
GTM channels
- Community
- API
- Docs
ICP
- Software developers
- Engineering teams
- Freelancers solopreneurs