M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
COMPARE UP TO 4 SERVERS

Compare MCP servers

Compare scores, permissions, risks, and fit in one decision-focused table.

DimensionDreamGraphA graph-governed architecture cognition layer for MCP-enabled software projects.MartinLoop MCP ServerAn independent control layer that gives AI coding agents budgets, brakes, and receipts
FMRS58 / 100 · C79 / 100 · B
Reliability8 / 2012 / 20
Security and permissions8 / 2018 / 20
Maintenance14 / 2016 / 20
Documentation16 / 2018 / 20
Setup experience12 / 2015 / 20
Best for
  • Developers whose codebases have grown beyond short-term memory, seeking AI that understands the system and explains changes.
  • Architects, system engineers, and product builders needing an architecture intelligence layer.
  • Projects using MCP clients like Claude Desktop or VS Code that benefit from persistent graph memory.
  • Engineering teams scaling AI coding agents while controlling spend
  • Platform teams that need auditable, rollback-aware agent execution records
  • Organizations using multiple agents (Claude, Codex, Gemini) under one governed flow
Not for
  • Users seeking a lightweight chat assistant rather than a complex knowledge graph system.
  • Projects preferring simple tooling without persistent state or architecture governance.
  • Environments without Node.js 20+ or a compatible shell.
  • Developers who just want a simple code-completion helper without run governance
  • Teams unwilling to accept local JSONL run records with HMAC signing
  • Environments without a supported adapter (Claude/Codex/Gemini CLIs, direct provider, or verifier-only)
Required permissions
  • Read project files to build the knowledge graph (scan, enrich).
  • Read and write instance configuration and database (SQLite by default, optional PostgreSQL).
  • Execute `dg` CLI commands to manage daemon lifecycle.
  • VS Code extension permissions: file access, editor integration, daemon connection.
  • Potential network access (via HTTP daemon) and LLM API or local model server connections.
  • Read/write access to the configured repo root (writable scope restrictable via --allow-path/--deny-path)
  • Ability to invoke the chosen agent CLI (Claude/Codex/Gemini, etc.) and its underlying model account
  • Local disk writes for run records, receipts, and share outputs (share/run-receipt., etc.)
  • Execution of user-configured verifier commands (e.g. npm test)
Risks and side effects
  • Cognitive outputs are advisory until backed by governed evidence or explicit review; expired or rejected ideas remain inspectable but not authoritative.
  • Ensure instance configuration is secured, especially when using HTTP transport; do not expose the daemon on untrusted networks.
  • Plugins (when enabled) must be trusted; untrusted plugins may pose risks, but mechanisms for gating and trust banners exist.
  • Backup data before major upgrades, as migration scripts may be needed for older versions (e.g., pre-v8.2.6).
  • Verifier commands are executed; poorly chosen verifier commands carry security risk despite preflight policy checks
  • Cost and token figures may be estimates (labeled with provenance) and should not be treated as settled accounting
  • Receipt integrity must be verified before use as trustworthy evidence; missing rollback evidence is flagged as EVIDENCE_BOUNDARY
  • Policy defaults come from martin.config.yaml and can be overridden by CLI flags; misconfiguration can loosen safety boundaries
Supported clientsClaude Desktop, VS CodeClaude Code, Codex, Gemini CLI, Cursor, VS Code, GitHub Copilot, Continue
Tools00