ClaudeHistoryMCP
Provides an MCP server plus optional CLI/hook to index and search Claude Code conversation history stored as JSONL under the local Claude projects directory, and to extract/surface prior decisions, solutions, and error-fix patterns. Optionally supports cloud sync via HTTP tools.
Score Breakdown
⚙ Agent Friendliness
🔒 Security
Local MCP execution appears unauthenticated and relies on local filesystem access; cloud sync uses an API key via environment variables but no details are provided about TLS enforcement, request signing, scope granularity, or safe logging. Cloud sync may transfer conversation content off-device, so access control and TLS details on the cloud server are important.
⚡ Reliability
Best When
You have substantial Claude Code history locally and want fast, tool-driven retrieval of prior decisions and fixes during development.
Avoid When
You cannot grant the process access to local Claude history files or you need a documented, production-grade API contract (schemas, error codes, pagination, SLAs) for automated agent use.
Use Cases
- • Search prior Claude Code sessions for fixes to recurring errors (e.g., ECONNREFUSED).
- • Retrieve summaries and structured context for a project based on recent sessions.
- • Discover recurring patterns/workflows from historical conversations.
- • Proactively inject prior project context at the start of new Claude Code sessions.
- • Sync indexed knowledge across devices/teams via a separate cloud server.
Not For
- • Auditing/compliance workflows that require formally verified summarization accuracy.
- • Use as a general-purpose web search or semantic retrieval service for non-Claude data.
- • High-assurance environments where local file access and indexing risks must be tightly controlled without additional hardening.
Interface
Authentication
No MCP auth details are provided in the README. Cloud sync appears to use an API key, but the cloud API’s auth model/scopes are not documented here.
Pricing
No pricing information stated; the project is MIT and described as a local MCP server with optional separate cloud server.
Agent Metadata
Known Gotchas
- ⚠ Tool behavior and parameter schemas are not fully specified in the README; agents may need to infer or inspect the MCP tool definitions at runtime.
- ⚠ Search/filtering relies on the specified filter syntax; malformed filters may reduce results or fail without documented guidance.
- ⚠ Local indexing depends on filesystem layout and incremental indexing/file-watcher behavior; stale indexes may occur if the watcher/index timing fails.
Alternatives
Full Evaluation Report
Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for ClaudeHistoryMCP.
AI-powered analysis · PDF + markdown · Delivered within 30 minutes
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
Scores are editorial opinions as of 2026-03-30.