In-Memoria

In Memoria is an on-machine (local-first) MCP server that learns from a codebase and exposes persistent “project blueprint” context, semantic/pattern search, file routing, and work-memory/pattern recommendations to AI coding assistants.

Evaluated Mar 30, 2026 (21d ago)
Repo ↗ DevTools ai-ml devtools mcp persistent-memory code-analysis semantic-search local-first rust sqlite surrealdb
⚙ Agent Friendliness
57
/ 100
Can an agent use this?
🔒 Security
23
/ 100
Is it safe for agents?
⚡ Reliability
31
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
85
Documentation
65
Error Messages
0
Auth Simplicity
95
Rate Limits
10

🔒 Security

TLS Enforcement
10
Auth Strength
20
Scope Granularity
0
Dep. Hygiene
55
Secret Handling
40

Local-first design reduces exposure over the network, but there is no documented authentication/authorization for MCP tool usage and no documented transport security needs because it’s invoked locally. Embeddings and learned data are stored locally (SQLite + SurrealDB with SurrealKV) but no details are provided about encryption at rest, access controls, or data deletion. Dependency list includes standard libraries; no vulnerability status is provided in the README/manifest. Agents should assume tool calls can access local codebase content and learned artifacts, so host/permissions hardening is important.

⚡ Reliability

Uptime/SLA
0
Version Stability
50
Breaking Changes
45
Error Recovery
30
AF Security Reliability

Best When

You want local-first, persistent coding assistance where an MCP-capable agent repeatedly queries learned context, search, and routing for the same repository over time.

Avoid When

You need a server with strong enterprise-grade access controls, documented rate limits, and guaranteed consistency/SLA; In Memoria is explicitly described as early-stage/WIP.

Use Cases

  • Persistent AI-assisted codebase navigation across sessions
  • Generating compact project context (tech stack, entry points, architecture) for an agent
  • Semantic/pattern-based code search to find relevant code by intent
  • Routed feature implementation suggestions (e.g., “password reset” → likely files)
  • Capturing and reusing architectural decisions and coding conventions (work memory/pattern prediction)

Not For

  • Multi-tenant hosted deployments without considering data isolation
  • Cases requiring a hosted HTTP API with standard auth and rate limits
  • Security-sensitive environments without auditing local storage and data handling
  • Use as a general knowledge base unrelated to a specific local codebase

Interface

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

Authentication

Methods: None documented for MCP server in the README (local invocation via npx/CLI)
OAuth: No Scopes: No

The README and manifest describe local-first MCP server usage via CLI. No authentication mechanism, API keys, or scope model is described for MCP tool calls.

Pricing

Free tier: No
Requires CC: No

Open-source (MIT) based on repository metadata; pricing not applicable beyond local environment costs.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • MCP tools are ignored unless the chat is in the agent/appropriate mode (not Ask/Edit)
  • MCP server must be running/configured (e.g., Claude Desktop/MCP config or VS Code .vscode/mcp.json)
  • Large codebases can be slow on first analysis; agents may time out or appear unresponsive during learning/ingestion
  • “Status: Work in Progress” and README notes documentation may be incomplete; tool behavior/edge cases may change

Alternatives

Full Evaluation Report

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

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