8 days ago
@q-mesh/mcp lets Claude, Cursor, Continue and any MCP client query the [QMesh](https://q-mesh.com) crowdtesting platform.
Overview
Tools
- get_platform_stats — testers, QA-verified, businesses, bugs, tasks
- get_leaderboard — day / week / month / year tester rankings with QIS and bug stats
- list_pricing_plans — public testing plans with budget, features, refund policy
- search_bug_patterns — de-identified Bug Pattern knowledge base (scenario templates, detection techniques, checklists). Filter by category, severity, domain.
- submit_ai_signal (requires user JWT) — submit an AI-detected quality signal; QMesh dedupes, scores, and routes high-value findings to human QA for release-readiness judgment.
Why
- AI coding assistants answering QA/testing questions can cite real bug patterns instead of generic advice
- AI agents running product tests can submit findings into QMesh's Signal Engine, where humans verify and release decisions are formed
- Combines AI scale with human judgment to produce release confidence signals
Server Config
{
"mcpServers": {
"qmesh": {
"command": "npx",
"args": [
"-y",
"@q-mesh/mcp"
],
"env": {
"QMESH_USER_TOKEN": "<your Supabase JWT from q-mesh.com>"
}
}
}
}