Memory Fabric MCP server
File-first, local-first MCP memory for AI coding assistants: Markdown + YAML, no vector DB.
2 stars212 downloads/wk
Reviews
Write oneNobody has reviewed Memory Fabric yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
Memory Fabric tools (21, 8 write)
write = sends, deletes, buys or postsRead from the package source without running it. The installed server may list more.
apply_dream_results_toolApply consolidation results generated by the client-agent's LLM run.
context_for_task_toolReturn a task-scoped memory pack: steering + top-k ranked store entries.
delete_memory_store_toolwrite actionRemove a memory-store file by its semantic path.
dream_toolwrite actionRun the Dreaming consolidation pass over memory files.
drop_review_toolwrite actionDrop a pending capture from the review queue (marks it dropped).
evaluate_dream_quality_toolScore the quality improvement produced by a Dream run.
evaluate_memory_fabric_toolEvaluate the overall quality of the Memory Fabric for this project.
initialize_memory_fabric_toolBootstrap Memory Fabric in the project at *cwd*.
keyword_search_toolSearch all memory files by keyword and return ranked results.
list_memory_store_toolList files in the memory store, optionally filtered by prefix and/or tags.
list_pending_reviews_toolList ``review_status: pending`` / ``needs-review`` captures waiting to be promoted.
prepare_dream_payload_toolPrepare prompt and payload for client-agent to perform consolidation.
promote_review_toolPromote a pending capture into architecture/, decisions/, or failures/.
propose_memory_patch_toolwrite actionPreview a proposed memory update as a unified diff without writing to disk.
read_combined_context_toolLoad and assemble the full memory context for the project.
read_memory_store_toolRead a single memory-store file by its semantic path.
read_section_toolRead a single flat memory section file by name.
write_failure_memory_toolwrite actionRecord an error -> fix pair. Call this immediately after fixing a bug.
write_local_memory_toolwrite actionWrite or append content to a flat steering/directive section file.
write_memory_store_toolwrite actionWrite a memory file to a semantic store path (e.g. 'architecture/decisions/auth-service').
write_session_journal_toolwrite actionAppend a timestamped session journal to the episodic memory store.
Public scan report
scanner v0.1.9 · 2026-09-20 · same rubric, same numbers if you re-run it
- Code scan54 source files scanned25/25
- –Live reliabilityno gateway calls yet and no remote to proben/a
- –Tool poisoningtools not inspected (local package is not executed); not countedn/a
- Auth qualitylocal package, no credentials required12/15
- Maintenancelast push 15 days ago15/15
- Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Install directly
claude mcp add memory-fabric -- uvx memory-fabric
Memory Fabric: common questions
- Is Memory Fabric MCP server safe?
- Yes, by our scan: it is graded A (92/100). Read the Memory Fabric safety report
- How do I install Memory Fabric?
- It runs on your machine. Copy the Claude Code, Claude Desktop or Cursor config from the install section.
- Does Memory Fabric need an API key?
- Not as far as the registry entry and our scan can tell: no credentials are declared or required.
- Is Memory Fabric maintained?
- The last commit was 15 days ago (2026-09-05). The latest release is v1.4.1.
- What can I use instead of Memory Fabric?
- Servers from other publishers that do the same job: IWE MCP server, Total Agent Memory MCP server and Projectmem MCP server. Compare all Memory Fabric alternatives.
Alternatives to Memory Fabric
Same job from other publishers: the closest match first, then the best rated.
- IWEMarkdown knowledge base as agent memory. Runs against the notes directory it is started in.not reviewedEstablishedA
- Total Agent MemoryPersistent local memory for coding agents: temporal knowledge graph, procedural and episodic recallnot reviewedEstablishedA
- ProjectmemCoding agent memory — one local MCP server for every project. Warns before repeating failed fixes.not reviewedGrowingA
- adrkit decision memoryDeterministic, offline, read-only ADR decision memory for coding agents. No model or network calls.not reviewedGrowingA
- rag-ratRepository intelligence, code graph, history, papertrail, and cross-agent memory for coding agents.not reviewedGrowingB