Engram
Related Servers
Alternatives to Engram
- AlicenseAqualityDmaintenanceThis MCP server provides persistent memory integration for chat applications by utilizing a local knowledge graph to remember user information across interactions.9374,489 npm6MIT

Mem0 MCP Serverofficial
AlicenseAqualityFmaintenanceAn MCP server that integrates with mem0.ai to help users store, retrieve, and search coding preferences for more consistent programming practices.9662Apache 2.0
Related Servers
- FlicenseNot gradedqualityAmaintenanceEnables MCP-compatible coding tools to share a portable, local-first AI work identity and memory layer, so preferences, decisions, lessons, and project context persist across sessions and clients. Users can inspect, edit, approve, and override what the AI receives, keeping the store on their own machine without a cloud account.-
- FlicenseNot gradedqualityAmaintenanceProvides a local-first, user-owned identity layer that lets MCP-compatible coding agents start from the same approved preferences, quality bar, lessons, and project context. Memory stays in inspectable local files that can be reviewed, edited, overridden, and carried across different AI coding tools.-
- AlicenseNot gradedqualityCmaintenanceProvides a local-first persistent memory layer for AI coding agents through MCP, enabling them to save, search, retrieve, update, and consolidate project context, technical decisions, debugging solutions, and developer preferences across sessions.MIT
- AlicenseNot gradedqualityBmaintenanceProvides a local-first persistent memory layer for coding assistants, enabling cross-project user preferences, per-project durable and working memory, session archives, and reflective recommendations through MCP tools.43 npmMIT
- FlicenseNot gradedqualityAmaintenanceGives MCP-compatible coding tools a shared, user-owned identity and knowledge layer stored in local files, so preferences, quality standards, lessons, decisions, and project context carry across tools and sessions. Everything is inspectable, editable, and overridable, with AI-proposed writes reviewed or auditable and a preview showing exactly what any caller would receive.-
- FlicenseNot gradedqualityAmaintenanceEnables a local-first, user-owned AI work identity layer that lets multiple MCP coding tools share approved preferences, quality standards, lessons, decisions, and project context, with local files you can inspect, edit, and override.-
TDQS
Scored across 19 tools
Write side is clearly split (add_lesson/add_decision/add_playbook vs memory_store routing), but the read side is crowded: get_user_context, get_resume_brief, get_recall, get_relevant_knowledge, search_knowledge, get_recent_context, and get_daily_log all return overlapping memory bundles. The descriptions do offer 'use X when Y' guidance that helps, but an agent still has to reason carefully to pick the right aggregator for a cold start.
Nearly everything follows verb_noun snake_case (get_*, add_*, save_*, update_*, search_*), which is predictable. Minor deviations are 'doctor' (bare noun) and 'wrap_up_session' / 'get_wrap_up_session_status', which break the pattern slightly but stay readable.
19 tools is in the heavy end of the acceptable band; most serve a genuine lifecycle stage (startup, retrieval, writeback, session-end, diagnostics), but several aggregator reads (resume_brief, recall, user_context, relevant_knowledge) could plausibly be consolidated. It is defensible for a full memory system yet feels broad.
Covers startup context, keyword/path retrieval, all three knowledge write types, revision history, identity read/update, project snapshot read/write, session wrap-up with status polling, daily logs, and diagnostics — a full lifecycle. The main gap is the absence of any delete/forget/archive operation over MCP, which is deferred to the CLI owner workflow.