ContextKeep

ContextKeep is a standalone memory server for AI agents that exposes MCP tools to store, retrieve, search, and list persistent memories. It supports local (stdio) and remote transports (SSE) plus a web dashboard for managing memories stored on the host (SQLite-backed per changelog).

Evaluated Mar 30, 2026 (0d ago)

Homepage ↗ Repo ↗ Ai Ml ai-ml agent mcp memory python webui sqlite

⚙ Agent Friendliness

/ 100

Can an agent use this?

🔒 Security

/ 100

Is it safe for agents?

⚡ Reliability

/ 100

Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality

Documentation

Error Messages

Auth Simplicity

Rate Limits

🔒 Security

TLS Enforcement

Auth Strength

Scope Granularity

Dep. Hygiene

Secret Handling

README claims 100% local storage (good for data exposure), but does not document authentication/authorization, TLS requirements for SSE, or rate limiting. If exposed beyond localhost, transport endpoints should be firewalled/reverse-proxied with TLS and access controls.

⚡ Reliability

Uptime/SLA

Version Stability

Breaking Changes

Error Recovery

Best When

You run it yourself (local/homelab) and want an MCP-compatible long-term memory for one or a small number of trusted clients on the same machine/network.

Avoid When

You need robust authentication/authorization, rate limit policies, and structured error contracts for untrusted clients; also avoid exposing the web UI/MCP transport to untrusted networks without additional protections.

Use Cases

• Give AI coding assistants long-term project memory (decisions, preferences, snippets)
• Maintain searchable personal or team knowledge across agent sessions
• Use deterministic memory lookup via an index directory (list_all_memories -> retrieve_memory)
• Manage memories via a local web dashboard (CRUD, search, browsing views)
• Run as a homelab service for multiple clients on a local network

Not For

• High-security or multi-tenant environments where strong access control is required out of the box
• Workloads needing a documented, public REST/SDK API beyond MCP
• Use cases requiring guaranteed formal data retention policies or auditable compliance controls

Interface

REST API

GraphQL

gRPC

MCP Server

Yes ↗

SDK

Webhooks

Authentication

Methods: None documented for MCP transport (stdio/SSE/ssh described without auth controls).

OAuth: No Scopes: No

README emphasizes local storage/privacy but does not describe authentication, authorization, or per-client access controls for MCP server or web dashboard. Remote SSH/SSE are described as transport options, not as security mechanisms.

Pricing

Free tier: No

Requires CC: No

Open-source/self-hosted; no usage-based pricing described.

Agent Metadata

Pagination

none

Idempotent

True

Retry Guidance

Not documented

Known Gotchas

⚠ Recommended protocol: call list_all_memories() first, then retrieve_memory(exact_key); using search_memories() for key lookup may reduce determinism.
⚠ No explicit guidance on handling missing keys, conflicting tags, or empty search results is provided in the README.

Alternatives

memGPT (agent memory framework) LangChain memory / vector stores (e.g., Postgres+pgvector, Qdrant) LlamaIndex + persistent stores Self-hosted knowledge bases with search (e.g., Meilisearch + DB) and custom tooling Other MCP memory/context servers (general category)

Full Evaluation Report

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

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

Score Monitoring

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

Continuous monitoring

$3/mo

API endpoint ↗ Agent guide ↗ Report inaccuracy

Scores are editorial opinions as of 2026-03-30.