memento

Agentic AI is great, and a real help, until you ask it to pickup where you left off last week, on a project you haven't discussed since then.  Now you're guessing, rebuilding, adding real risk, and losing time.  Our memento plugin resolves this issue, and makes your AI agents think project first, and become true partners to your development staff.

• Project-first memory, active project context, and recall-before-asking behavior. Switch project context with simple chat commands - your agent now multitasks with you.
• Explicit agent tools for recall, writes, operational lookup, inspection, and maintenance.
• Postgres source-of-truth records with lifecycle, audit, links, mentions, read events, and encrypted secret payloads.* 
• Team/Project sharing and device sync, each developer adds to  the projects shared memory.* 
• Agent drift hardening through checked guardrails, smoke prompts, and repo-owned workflows.  Ideal when using multiple models for token aggregation, specific tooling, and cost control. 
• Obsidian-compatible export and review workflows for human inspection and graph visualization.*
• Portable onboarding context so project facts and operating preferences move between machines.*

*Available with Pro version, OS version uses local SQLite