MemoryMesh

MemoryMesh is a local knowledge-graph MCP server for persisting structured “memory” (nodes and edges) for AI models. It loads schema files to dynamically generate add/update/delete tools for each entity type, stores data in a local JSON memory file, and optionally provides a separate Memory Viewer web app for visualization.

Evaluated Mar 30, 2026 (22d ago)
Repo ↗ Ai Ml mcp knowledge-graph agent-tools structured-memory typescript local-first schemas
⚙ Agent Friendliness
59
/ 100
Can an agent use this?
🔒 Security
30
/ 100
Is it safe for agents?
⚡ Reliability
35
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
82
Documentation
70
Error Messages
0
Auth Simplicity
95
Rate Limits
0

🔒 Security

TLS Enforcement
50
Auth Strength
5
Scope Granularity
0
Dep. Hygiene
45
Secret Handling
60

Likely no auth/authorization controls are implemented or documented (local server). TLS is not applicable from the README (MCP server appears started locally via node). Security posture depends on how users deploy/expose the server; running locally reduces network risk but does not address tool-level constraints or access control. Dependency list is minimal in manifest but specific audit/CVE status is not provided.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
45
Error Recovery
40
AF Security Reliability

Best When

You want local, schema-driven structured memory for an AI client via MCP (e.g., Claude Desktop) and can manage data persistence on disk.

Avoid When

You need remote access across untrusted networks or fine-grained access control, or you require strong durability/concurrency guarantees beyond a single JSON file.

Use Cases

  • Text-based RPGs and interactive storytelling with persistent world state
  • Structured memory for agents that need nodes/relationships (e.g., NPCs, locations, artifacts)
  • Simulations such as social-network or organizational planning graphs
  • Building custom entity/relationship schemas and letting an MCP client call generated tools

Not For

  • Multi-tenant cloud deployments requiring managed hosting
  • High-security environments needing authentication/authorization controls in the server
  • Workloads needing relational database features, concurrency control, or large-scale storage beyond a JSON file

Interface

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

Authentication

Methods: None described (local MCP server started via node command)
OAuth: No Scopes: No

No authentication or authorization mechanisms are described in the provided README/manifest; access appears to be controlled implicitly by local process/network exposure.

Pricing

Free tier: No
Requires CC: No

Open-source (MIT) and appears local/self-hosted; no pricing details provided.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Schema recognition depends on filenames and naming conventions (schema name must start with add_).
  • Default schema/memory file locations are version-dependent (e.g., dist/data/schemas since v0.2.7).
  • Local persistence via JSON file implies concurrent writes could be problematic; no concurrency guarantees are stated.

Alternatives

Full Evaluation Report

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

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 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered