Provides self-hosted, versioned project memory with full-text and vector search, enabling search, retrieval, upsertion, and history management of project entries via MCP tools.
Durable, inspectable memory for MCP agents. Preserves decisions, preferences, and project knowledge across sessions with full provenance and version history.
A persistent, project-scoped memory layer for AI agents, supporting hybrid retrieval (vector, keyword, and tag matching) and sharing across different MCP clients like Claude Code, Qoder, or Cursor.
Provides persistent memory (L1/L2/L3 encrypted), web-search savings, and thread memory for AI agents, enabling token savings and context retention across MCP clients.
Provides persistent, scoped shared memory for collaborating AI agents, with tools for storing observations, semantic recall, and handoff workflows. Backed by PostgreSQL and exposed through MCP.