Each MCP server is run through the gauntlet: a live LLM agent attempts generated tasks using only the server's tools, alongside schema, description, security, reliability, and robustness checks. Grade is the weighted overall; a critical security finding caps it.
| # | Server | Grade | Score | Task success | Security | Tools | Scanned |
|---|---|---|---|---|---|---|---|
| 1 | cleanor-mcp | A | 100.0 | 100 | ✓ | 22 | 2026-07-29 |
| 2 | demomagic-mcp | A | 100.0 | 100 | ✓ | 18 | 2026-07-29 |
| 3 | kavel-mcp | A | 100.0 | 100 | ✓ | 4 | 2026-07-29 |
| 4 | pretrip-mcp | A | 100.0 | 100 | ✓ | 2 | 2026-07-29 |
| 5 | ssid-mcp | A | 100.0 | 100 | ✓ | 1 | 2026-07-29 |
| 6 | raven-mcp | A | 97.7 | 100 | ✓ | 105 | 2026-07-29 |
| 7 | testivai-mcp | B | 82.6 | 33 | ✓ | 8 | 2026-07-29 |
| 8 | spideriq-mcp-leads | C | 76.9 | 45 | ✓ | 51 | 2026-07-29 |
| 9 | nudg3-mcp | C | 75.7 | 7 | ✓ | 11 | 2026-07-29 |
| 10 | mcpcap | C | 73.0 | 0 | ✓ | 9 | 2026-07-29 |
| 11 | kawacode-mcp | D | 65.2 | 0 | ✓ | 26 | 2026-07-29 |
| 12 | shipeasy-mcp | D | 64.9 | 5 | ✓ | 115 | 2026-07-29 |
| 13 | ai-dossier-mcp-server | D | 63.1 | 0 | ✓ | 10 | 2026-07-29 |
Agent model: gemini:gemini-flash-latest · generated 2026-07-29T04:21+00:00 · scores from a live agent are stochastic (repeated and averaged); ⚠ = static tool-poisoning in a description (caps the grade), ⚡ = injection in a live tool output (does not cap).
Scored by mcp-gauntlet 0.7.0. Scoring changes between releases, so compare scores only within the same version — each row's Scanned date is when that score was measured, not when this page was rebuilt.
These servers were not measured the same way as the ranked ones — the agent either never scored them, or stopped partway when a tool hung. Either way their score rests on a different basis: the overall is a weighted mean over the dimensions present, over however many runs completed, so a missing Agent Task Success (the heaviest dimension) or a short sample inflates the number relative to the ranked table. Each row says which applied.
| # | Server | Grade | Score | Task success | Security | Tools | Scanned |
|---|---|---|---|---|---|---|---|
| — | aetherwealth-mcp | A | 97.4 | — | ✓ | 48 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | autonomad-travel | A | 99.8 | — | ✓ | 8 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | callmcp-server | A | 96.2 | — | ✓ | 14 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | justdrop-mcp | A | 100.0 | — | ✓ | 4 | 2026-07-29 |
| no read-only tools to test (all excluded as possibly-mutating) | |||||||
| — | percher-mcp | A | 97.2 | — | ✓ | 51 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | popdot-mcp | A | 100.0 | — | ✓ | 5 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | radmail-mcp | A | 96.5 | — | ✓ | 12 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | rapay-mcp-server | A | 100.0 | — | ✓ | 7 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | spideriq-mcp-gate | A | 99.8 | — | ✓ | 19 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | synchronex-mcp-proxy | A | 96.6 | — | ✓ | 7 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| — | thrain-blackout-mcp | A | 100.0 | — | ✓ | 2 | 2026-07-29 |
| no read-only tools to test (all excluded as possibly-mutating) | |||||||
| — | thrain-spotlight-mcp | A | 100.0 | — | ✓ | 1 | 2026-07-29 |
| no read-only tools to test (all excluded as possibly-mutating) | |||||||
| — | vaaya-mcp | A | 100.0 | — | ✓ | 11 | 2026-07-29 |
| no agent evaluation (static checks only) | |||||||
| # | Server | Grade | Score | Task success | Security | Tools | Scanned |
|---|---|---|---|---|---|---|---|
| — | 3dstreet-mcp | exposes no tools | |||||
| — | adeu-mcp-server | could not evaluate: timed out after 120s | |||||
| — | agentutility-mcp-agentops | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: [@agentutility/mcp-agentops] FATAL: X402_PRIVATE_KEY env var is required. / Set it to a hex-encoded EVM private key with USDC balance | |||||
| — | agentutility-mcp-bestiary | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: [@agentutility/mcp-bestiary] FATAL: X402_PRIVATE_KEY env var is required. / Set it to a hex-encoded EVM private key with USDC balance | |||||
| — | ankimcp-anki-mcp-server | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: [31m✗[0m Failed to start MCP HTTP server: listen EADDRINUSE: address already in use 127.0.0.1:3000 | |||||
| — | bourdon | could not evaluate: McpError: Connection closed | |||||
| — | businys-mcp-server | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: npm error code EUNSUPPORTEDPROTOCOL / npm error Unsupported URL Type "workspace:": workspace:* | |||||
| — | clize-clize | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: build 做一个作品(端到端):build site = 站点/UI · build clip = / 短片;方法/校验免费,成品步骤 💰 / help [command] display | |||||
| — | contextlayer-mcp | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: npm error could not determine executable to run | |||||
| — | contextstudios-mcp | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: npm error could not determine executable to run | |||||
| — | contio-mcp | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: npm error could not determine executable to run | |||||
| — | ctxl | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: npm error could not determine executable to run | |||||
| — | engrava-mcp | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: raise StoreResolutionError(msg) / | engrava_mcp.config.StoreResolutionError: No engrava store configured. Set ENGRAVA_MCP_CONFIG to | |||||
| — | meetlark-mcp-server | could not evaluate: timed out after 120s | |||||
| — | nauro | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: Installed 46 packages in 782ms | |||||
| — | oobe-protocol-labs-sap-mcp-server | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: requireStack: [ '~/workplace/git/mcp-gauntlet/package.json' ] / } / Node.js v22.23.1 | |||||
| — | perplexity-ai-mcp-server | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: Error: PERPLEXITY_API_KEY environment variable is required | |||||
| — | positif-ai | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: ╰─▶ Because there is no version of positif-ai==0.2.1 and you require / positif-ai==0.2.1, we can conclude that your requirements are | |||||
| — | reka-mcp | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: File "~/.cache/uv/archive-v0/ogNjuR7BrWL7Q2Ed/lib/python3.14/site-packages/reka_mcp/server.py", line 10, in <module> / from mcp.serve | |||||
| — | snowsure-mcp-server | could not evaluate: timed out after 120s | |||||
| — | telbase | could not evaluate: McpError: Connection closed | |||||
| — | toolprint-hypertool-mcp | could not evaluate: McpError: Connection closed | |||||
| — | trusty-squire-mcp | could not evaluate: MCPConnectionError: McpError: Connection closed — the server said: ebsockify / cloudflared is not in Debian/Ubuntu repos — install its package: curl -L https://github.com/cloudflare/cloudflared/releas | |||||
| — | unpaper-cli | could not evaluate: McpError: Connection closed | |||||
Listed here? Paste this into your README — it reads this board's published score and updates whenever the board is regenerated. Swap cleanor-mcp for your server's slug: the filename of its row link, without the .html.
[](https://ghalebdweikat.github.io/mcp-gauntlet/survey/)
Not listed, or think a score is wrong? Open an issue on the repository — re-runs and corrections are free. How these scores are computed, what they are not, and the disclosure policy behind publishing them: see METHODOLOGY.md.