Qai Consultant MCP
Keyless local QA server for standards lookup, effort estimates, doc review, test analysis
- Pricing
- Open source
- Platforms
- CLI, Linux, macOS
- Runs
- Locally
- Version
- 3.9.3
- Listing updated
- today
Similar MCP servers
About Qai Consultant MCP
Qai Consultant MCP is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.
The MCP server is a local companion to an open-source agent that acts like a senior QA architect. The README says it needs no API keys and no Pinecone account, runs with uvx, and lets Claude Code, Claude Desktop and other clients ground their QA work in the same standards and numbers. The registry entry lists standards retrieval, effort estimation, document review and test analysis.
The wider project covers generating a risk register, a PERT-based effort estimation report, an ISTQB-aligned test strategy and an IEEE 829-aligned test plan from an 11-question dialogue. It also has deterministic modes that score an existing test plan from 0 to 100 across six dimensions, analyse JUnit XML or CSV results for flaky and slow tests, and estimate a TMMi maturity level with an EU AI Act readiness check.
Its knowledge base draws on ISTQB syllabuses, OWASP, ISO 26262, A-SPICE, IEEE 829, ISO/IEC 25010 and the EU AI Act. The full agent needs Mistral, OpenRouter and Pinecone keys; the MCP package does not.
Install
Run in a terminal
claude mcp add qai-consultant-mcp -- uvx qai-consultant-mcpAdd to claude_desktop_config.json
{
"mcpServers": {
"qai-consultant-mcp": {
"command": "uvx",
"args": [
"qai-consultant-mcp"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"qai-consultant-mcp": {
"command": "uvx",
"args": [
"qai-consultant-mcp"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"qai-consultant-mcp": {
"type": "stdio",
"command": "uvx",
"args": [
"qai-consultant-mcp"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"qai-consultant-mcp": {
"command": "uvx",
"args": [
"qai-consultant-mcp"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"qai-consultant-mcp": {
"command": "uvx",
"args": [
"qai-consultant-mcp"
]
}
}
}Using several servers? Build one config file with the MCP config generator.
What it does
- Standards retrieval
- Looks up QA standards such as ISTQB, OWASP and IEEE 829.
- Effort estimation
- PERT-based effort breakdown with confidence score.
- Document review
- Scores test plans and strategies across ISTQB and IEEE 829 dimensions.
- Test results analysis
- Finds flaky, never-run and slowest tests from JUnit XML or CSV.
- Maturity assessment
- Indicative TMMi level and EU AI Act readiness check.
- Keyless local run
- Runs through uvx with no API keys.
Good for
- Estimate QA effort for a new project
- Review a draft test plan for gaps
- Find flaky tests in CI results
- Draft a risk register grounded in standards
- Check QA maturity of a team process
Pricing
Open source
Open source; MCP server needs no API keys
Qai Consultant MCP FAQ
Does the MCP server need API keys?
No, it is keyless and local.
How do I run it?
With uvx qai-consultant-mcp, or install qai-consultant-mcp from PyPI.
Is there a live demo?
Yes, the README links a hosted Streamlit app for the full agent.
What standards does it use?
ISTQB, OWASP, ISO 26262, A-SPICE, IEEE 829, ISO/IEC 25010 and the EU AI Act, among others.
Is Qai Consultant MCP free?
Yes. Qai Consultant MCP is open source, so you can use and self-host it for free.
How do I add Qai Consultant MCP to Claude Code?
Run this command in your terminal: claude mcp add qai-consultant-mcp -- uvx qai-consultant-mcp
What are the best alternatives to Qai Consultant MCP?
The closest alternatives to Qai Consultant MCP are REA, Context7, Chrome DevTools MCP, Codebase Memory and GitHub.
Where this comes from
- Description: written from the project's README.
- Install steps: from the official MCP Registry entry.


