Skip to content
Agentmart
Submit a tool

Local Memory

Persistent local AI memory.

Connector by studiomeyer-io · GitHub account verified by the MCP Registry

IndexedMITLocal (stdio)★ 15
Write a review

Install

Works with Claude Code, Codex, VS Code, Cursor, Windsurf, Claude Desktop. Read the repository before you give any tool access to your data.

npx agentmarket install connector/studiomeyer-io-local-memory-mcp

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Persistent local AI memory. SQLite + FTS5 + vector search, knowledge graph. No cloud, no API keys.

Use cases: agent-memory

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this server can do and what it asks for, from its tool list, install config and code. Checked Oct 25, 2026.

Tools
8 read · 8 write · 2 delete · 7 not yet classified
Secrets
None asked for
Install scripts
None found
Runs as
Node.js package

Tools by what they do 25

Read from the source code; the running server may list different tools.

Read 8 · Look things up without changing anything.

  • memory_contradictions
  • memory_entity_search
  • memory_goal
  • memory_health
  • memory_insights
  • memory_profile
  • memory_recall
  • memory_search

Write 8 · Create, change, send or run things.

  • memory_decide
  • memory_entity_create
  • memory_entity_observe
  • memory_entity_relate
  • memory_import
  • memory_learn
  • memory_learn_update
  • memory_session_start

Delete 2 · Delete, drop, revoke or wipe things.

  • memory_entity_delete
  • memory_learn_archive

Not yet classified 7 · We could not tell from the name and description.

  • memory_entity_open
  • memory_export
  • memory_guide
  • memory_learn_bulk
  • memory_observation_supersede
  • memory_reflect
  • memory_session_end

Secrets and settings it asks for

Other settings (environment variables or headers).

  • MEMORY_DB_PATH
  • MEMORY_EMBED_CACHE_DIR
  • MEMORY_EMBED_DISABLED
  • MEMORY_EMBED_MODEL

Network hosts it talks to 2

  • github.com · found in code
  • studiomeyer.io · found in code

Files it touches

  • Your home directory (~)
  • Any path the agent passes to its tools

Tools (25)

Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].

  • memory_contradictions

    Scan observation pairs with high cosine similarity but disagreeing negation/confidence. LLM-free heuristic. Requires sqlite-vec. Per-entity observations are capped (maxObservationsPerEntity, default 200) to bound the O(N^2) self-join.

  • memory_decide

    Record a decision with reasoning and alternatives.

  • memory_entity_create

    Create an entity explicitly (without an initial observation). Idempotent on name+entityType. Embeds name+summary so vector/hybrid search over entities finds it.

  • memory_entity_delete

    Delete an entity and all its observations + relations. Destructive — use with care.

  • memory_entity_observe

    Record a fact about an entity. Creates the entity if missing.

  • memory_entity_open

    Load an entity with all its current observations and relations. Pass `asOf` (ISO date) for a bi-temporal point-in-time view.

  • memory_entity_relate

    Create a typed relation (edge) between two entities.

  • memory_entity_search

    Search entities by name or observation content (FTS5 fuzzy).

  • memory_export

    Export the whole memory to a versioned JSON envelope (learnings, decisions, knowledge graph, sessions, profile). Embeddings are re-derived on import. Same envelope imports into memory.studiomeyer.io (hosted tier).

  • memory_goal

    Read, set, or clear a single user goal stored in the profile table.

  • memory_guide

    On-demand help. Topics: quickstart, session, search, entities, learn, privacy.

  • memory_health

    SQLite integrity check + page-count + DB-size + WAL status. No input.

  • memory_import

    Import a memory_export envelope. Purely additive + idempotent (INSERT OR IGNORE on source ids), FK-safe, re-embeds on the fly. Pass {data: <envelope>}.

  • memory_insights

    Stats and reflection: days of memory, totals, category breakdown, entity type breakdown.

  • memory_learn

    Store a learning (pattern, mistake, insight, etc.). Gatekeeper prevents duplicates.

Show 10 more tools
  • memory_learn_archive

    Soft-delete a learning. Row stays for asOf-style queries; never resurfaces in recall/search. Optional reason.

  • memory_learn_bulk

    Batch-insert many learnings in one atomic call (parallel embed, exact-duplicate skip). For restores, migrations, seeding.

  • memory_learn_update

    Edit a live learning (content, confidence, tags). Re-embeds atomically when content changes.

  • memory_observation_supersede

    Retire an observation by setting valid_to (tombstone). Pairs with memory_contradictions. Pass supersededById to use the newer fact's valid_from as the cutoff (Zep fact-supersession). Row stays for asOf queries; drops out of live search/entity views.

  • memory_profile

    Read or write a user profile field (name, role, preferences, …). Stored locally.

  • memory_recall

    Quick FTS5 keyword recall over learnings only (no vector/hybrid — use memory_search for that). Optional project/tags scoping. Omit query for most-recent.

  • memory_reflect

    LLM-free reflection across recent memory: most-used + stale learnings, hot entities, open decisions. Returns structured data + markdown.

  • memory_search

    Unified hybrid search across learnings, decisions, entities, and observations (FTS5 BM25 + vector cosine fused with RRF). mode: fts|vector|hybrid (default hybrid). Optional project/tags scoping.

  • memory_session_end

    End the current session with an optional summary. Call at the end of each conversation.

  • memory_session_start

    Start a session. Loads context from previous sessions. Call this FIRST in every conversation.

Signals

Stars
15
Forks
6
Open issues
0
Last commit
Oct 21, 2026 (9 days ago)
Last release
Aug 19, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

Registry name
io.github.studiomeyer-io/local-memory-mcp
Version
2.4.3
Auth
No auth
Packages
npm @studiomeyer/local-memory-mcp
License
MIT
Repository
github.com/studiomeyer-io/local-memory-mcp
First seen
Oct 24, 2026
Last checked
Oct 24, 2026
Source
Official MCP Registrygithub-graphql
Cite this page

Local Memory MCP server. Agent Mart. https://agentmart.dev/connectors/studiomeyer-io-local-memory-mcp (last updated 2026-10-24)