memsearch

Markdown-first memory system for AI agents that stores memories as human-readable .md files with vector-based semantic retrieval via Milvus. Features smart deduplication, live file sync, hybrid search (dense + BM25), and a Claude Code plugin.

Evaluated Mar 08, 2026 (0d ago) vlatest
Homepage ↗ Repo ↗ AI & Machine Learning memory semantic-search rag milvus embeddings claude-code agent-memory markdown vector-database deduplication
⚙ Agent Friendliness
60
/ 100
Can an agent use this?
🔒 Security
55
/ 100
Is it safe for agents?
⚡ Reliability
29
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
68
Documentation
75
Error Messages
50
Auth Simplicity
72
Rate Limits
40

🔒 Security

TLS Enforcement
60
Auth Strength
55
Scope Granularity
35
Dep. Hygiene
68
Secret Handling
60

Local-first architecture. API keys for embedding providers stored in config. Markdown files are plaintext (no encryption at rest). Milvus connection may or may not use TLS depending on deployment.

⚡ Reliability

Uptime/SLA
0
Version Stability
35
Breaking Changes
30
Error Recovery
50
AF Security Reliability

Best When

You want persistent, searchable memory for AI agents that stores everything as human-readable markdown files you can version control.

Avoid When

You need a production-scale vector database (use Milvus directly), don't want to manage Milvus, or need non-markdown storage.

Use Cases

  • Persistent memory for AI coding agents across sessions
  • Semantic search over project documentation and notes
  • RAG pipeline with git-friendly markdown storage
  • Claude Code plugin for automatic memory persistence

Not For

  • High-throughput real-time search (latency-sensitive production systems)
  • Teams not comfortable running Milvus infrastructure
  • Simple key-value storage needs
  • Non-AI applications without embedding requirements

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
Yes
Webhooks
No

Authentication

Methods: api_key
OAuth: No Scopes: No

Requires API key for chosen embedding provider (OpenAI, Google, Voyage, etc.). Milvus connection string for remote instances. No auth for local usage with local embeddings.

Pricing

Model: free
Free tier: Yes
Requires CC: No

Open source (MIT). Free to use. Embedding API costs depend on chosen provider. Milvus can run locally for free or via Zilliz Cloud (paid).

Agent Metadata

Idempotent
True
Retry Guidance
Not documented

Known Gotchas

  • Requires Milvus running (local or remote) -- adds infrastructure dependency
  • Python 3.10+ required
  • Embedding provider costs can add up with large memory stores
  • New project (Feb 2026) -- API surface may still be evolving
  • File watcher must be running for live sync to work

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for memsearch.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-08.

6961
Packages Evaluated
25669
Need Evaluation
173
Need Re-evaluation
Community Powered