Score breakdown
Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.
Why it matters
Useful for developer teams using AI coding agents who want a persistent memory layer that survives session restarts, model upgrades, and agent switches (the one-SQLite-file design is the headline differentiator — no cloud, no API keys, no Docker, no vector database): PMB is the Apache-2.0 local-first persistent memory for AI coding agents (Claude Code / Cursor / Codex) with 29 MCP tools and a live
Who should use it
Who should skip it
Skip oleksiijko/pmb unless the captured evidence suggests it solves a problem you are actively working on.
About this signal
oleksiijko/pmb is tracked by RepoRadar as a local-first persistent memory fo in the Apache-2.0 local-first persistent memory for AI section. It was first seen on 2026-06-25 and last updated on 2026-06-25. The current verdict is 'try now' with a Silver tier and easy setup difficulty. oleksiijko/pmb leads on setup ease (8.8) and workflow potential (8.6); its lowest signal is momentum (6.0), so factor that in before investing setup time. This page summarizes the evidence RepoRadar has captured from captured source metadata. The score, tier, risk label, and verdict on this page are never influenced by sponsorship, ads, or tips — they reflect only the usefulness, popularity, novelty, momentum, maturity, and evidence signals described in the RepoRadar methodology.
How this item is evaluated
RepoRadar assigned oleksiijko/pmb a composite score of 7.5 out of 10, placing it in the Silver tier. This score combines weighted sub-signals: usefulness (35%), novelty (18%), momentum (14%), maturity (10%), open-source/build quality (7%), evidence quality (6%), workflow potential (6%), and setup ease (4%). Popularity is tracked separately at 85.0 and never affects the composite score or tier. The risk label of 'low' reflects inherent user-impacting hazards, not generic novelty. Items with no risk flag may still require normal code review before production use.
Putting this into practice? Read How to evaluate an AI tool before you adopt it for the checklist behind this score.
Risk explanation
**The SQLite file is the single point of failure — back it up.** The design is intentionally one SQLite file on disk, which is the right design for portability and ownership, but it also means the user's home directory backup is the only durable copy. For any production adoption, set up an automatic backup of the SQLite file (a cron job, a versioned copy in git, or a backup tool) before relying on PMB for a workflow that depends on the memory surviving; **85 stars and an early-stage project — verify the 29 MCP tools' coverage on the team's target use case.** PMB is at 85 stars with last push 2026-06-25, and the project's breadth (29 tools covering decisions + lessons + project facts + PDFs + personal facts) is the headline differentiator vs. a single-purpose memory layer. For any production adoption, run a representative workflow against the team's existing Claude Code / Cursor / Codex setup and confirm the 29 tools cover the team's specific use case (a gap in tool coverage would surface as the agent failing to recall a class of memory); **No multi-user or multi-tenant model — the SQLite file is single-user.** The design is intentionally single-user, single-machine. For teams that want shared memory across multiple developers, PMB's design is a starting point (the SQLite file is shareable as a file) but not a multi-tenant memory system. For multi-user or multi-tenant workflows, evaluate the team's existing collaboration tool and confirm PMB's single-user design matches the team's access-control model.