qai-consultant-mcp

Listed on the official MCP registry (io.github.gvasile29/qai-consultant-mcp), Glama, and Awesome MCP Servers.

A local, fully keyless MCP server: standards-grounded QA knowledge retrieval (ISTQB, OWASP, IEEE, ISO, EU AI Act), deterministic QA effort estimation, QA document quality review, and test-results health analysis — callable directly from Claude Code, Claude Desktop, or claude.ai.

No API keys, no Pinecone, no cloud LLM calls. It runs a local embedding index over a self-authored QA knowledge base and does the estimation math itself; the client LLM writes the narrative, this server just supplies grounding and numbers.

This package is the MCP companion to QAI Consultant, an AI QA Architect web app / CLI. If you're looking for the full app (Test Strategy / Risk Register / Effort Report generation with a browser UI), see the main project instead — this package is just the MCP server piece of it.

Install

uvx qai-consultant-mcp

Claude Code:

claude mcp add qai-consultant -- uvx qai-consultant-mcp

Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "qai-consultant": {
      "command": "uvx",
      "args": ["qai-consultant-mcp"]
    }
  }
}

First run downloads the embedding model (sentence-transformers/all-MiniLM-L6-v2, CPU-only) and builds a local index — this takes a minute or two the first time, then it's cached.

Tools

Prompts

• qa_project_interview — the project-intake interview (11 questions covering scope, tech stack, team, timeline, risks, compliance)
• risk_register_structure — Risk Register document structure + grounding instructions
• test_strategy_structure — Test Strategy document structure + grounding instructions
• test_plan_structure — IEEE 829-aligned Test Plan structure + grounding instructions

Each *_structure prompt instructs the client to ground its generation in retrieve_qa_knowledge results with [Source N] citations, and to label the output as AI-generated.

Privacy

Usage telemetry is off by default. Set QAI_TELEMETRY=1 to opt in. Even then, only the tool name, a success flag, duration, retrieval k/category, package/Python version, OS family, and a random anonymous install ID are sent — never your query text, project details, or knowledge-base content.

Source

github.com/gvasile29/qai-consultant — Apache 2.0 licensed.