MCP QUEEN
io.github.invertedhoologan/eval-engine-mcp A
Pay-per-call AI evaluation MCP server. Score LLM outputs against benchmark rubrics via Workers AI.
AI & Agentsv0.1.0streamable-httpeval.zuluworksai.com
Grade evidence — probed 2026-07-17 17:48 UTC
| Criterion | Points | Observed |
|---|---|---|
| reachability | 25 / 25 | HTTP 200, initialize accepted in 185ms |
| protocol | 15 / 15 | valid JSON-RPC initialize result, protocolVersion 2024-11-05, serverInfo eval-engine-api@0.1.0 |
| tooling | 35 / 35 | tools/list OK: 3 tools e.g. "evaluate"; 3/3 described, 3/3 fully-typed schemas, median description 214 chars |
| latency | 10 / 10 | initialize round-trip 185ms |
| provenance | 5 / 15 | description present; no repository URL; version 0.1.0; namespace io.github.invertedhoologan does NOT match remote host (expected *.invertedhoologan.github.io) |
Score 90/100 · latency 185ms · 3 tools · protocol access: open ⓘ
Trust receipt
Security, data integrity, citation quality, and advertised claims are measured separately from protocol uptime. Missing evidence is marked unverified, never converted into a pass.
| Dimension | Metric | Status | Evidence | Source |
|---|---|---|---|---|
| security | authentication_boundary | observed | Live protocol probe observed auth_state=open. This describes discovery/access behavior; it is not by itself a vulnerability finding. Observed value: open | live_probe |
Machine-readable receipt: /api/trust/io.github.invertedhoologan/eval-engine-mcp.json
Connect your agent
Claude Code:
claude mcp add --transport http eval-engine-mcp https://eval.zuluworksai.com/mcp
Generic MCP client config:
{ "mcpServers": { "eval-engine-mcp": { "type": "http", "url": "https://eval.zuluworksai.com/mcp" } } }
MCP Queen is a graded index, not a middleman — your agent connects directly to the server above. Check the grade and evidence first; that's the point.
Share this server: permalink https://mcpqueen.com/s/io.github.invertedhoologan/eval-engine-mcp · referral link https://mcpqueen.com/go/io.github.invertedhoologan/eval-engine-mcp (counts as “routed via the queen”).
Own this server? Embed your grade badge
Live operational-grade badge, re-probed continuously — not a security or data-quality certification. Put it in your README and link to the complete receipt:
[](https://mcpqueen.com/s/io.github.invertedhoologan/eval-engine-mcp)
Think the grade is wrong? Fix the finding the evidence shows, then the next probe cycle picks it up automatically (full cycle ≈ 3 days) — or open a dispute via the MCP endpoint.
Queen Watch
Email alerts when the grade changes or the endpoint stops answering. Double-opt-in, one-click unwatch, free while in beta.
Probe history
| When (UTC) | Grade | Score | Latency |
|---|---|---|---|
| 2026-07-17 17:48 | A | 90 | 185ms |