Skip to main content
Glama
Quality Transformation Coach Assistent MCP server

Quality Transformation Coach Assistent MCP server

Quality Transformation Coach MCP Server

Een MCP-server die als assistent dient voor software quality transformation — bedoeld voor zowel de quality transformation coach (overzicht, advisering, maturity assessments) als het development/test team (hands-on kwaliteitsanalyse, testondersteuning, CI/CD-inzichten).

Functionaliteiten

De server biedt 16 tools verdeeld over 5 categorieën:

Code Analyse

Tool

Beschrijving

analyze_code_quality

Scan een repo voor code kwaliteits-metrieken (complexity, duplication, smells)

analyze_test_coverage

Analyseer test coverage en identificeer gaps

detect_test_patterns

Herken test patterns (unit/integration/e2e) en anti-patterns

analyze_flaky_tests

Detecteer flaky tests op basis van CI-logs

Issue & Defect Analyse

Tool

Beschrijving

defect_trend_analysis

Analyseer issue-trends (aanwas, resolutietijd, severity)

quality_hotspot_detection

Identificeer modules met hoogste defect-dichtheid

root_cause_categories

Categoriseer bugs per oorzaak (code, design, requirements, security)

CI/CD

Tool

Beschrijving

pipeline_health

Analyseer CI/CD pipeline gezondheid (success rate, doorlooptijd)

test_result_summary

Agregeer test results vanuit CI runs

cicd_readiness_scan

Voer een CI/CD Readiness Scan uit (102 vragen, 5 domeinen)

quality_gate_check

Controleer of kwaliteits gates behaald zijn

AI Readiness

Tool

Beschrijving

ai_readiness_scan

Voer een AI Readiness Scan uit (60 vragen, 5 AI-domeinen)

Kennis & Maturity

Tool

Beschrijving

maturity_assessment

Voer een TMMi/ISO 25010 maturity scan uit

quality_recommendations

Geef context-gevoelige verbeteradviezen

framework_lookup

Raadpleeg kennisbank voor kwaliteitsmodellen

generate_quality_report

Genereer een compleet kwaliteitsrapport in Markdown

Related MCP server: MCP Test Failure Analysis Server

Installatie

Vereisten

  • Python 3.11+

  • GitHub Personal Access Token (voor issue/PR data)

Lokale installatie

git clone git@github.com:Cerios-TechLab/quality-transformation-coach-assistent.git
cd quality-transformation-coach-assistent
python3.11 -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"

Configuratie

Kopieer config.example.yaml naar config.yaml en pas aan:

github:
  token_env: GITHUB_TOKEN
  default_org: "Cerios-TechLab"

cicd:
  provider: github_actions

coverage:
  source: local
  format: cobertura

projects:
  - name: "cerios-clinic"
    repo: "/srv/cerios-clinic"
    github: "Steavy/cerios-clinic"

OpenCode integratie

Voeg toe aan ~/.config/opencode/opencode.json onder mcp:

"quality-coach": {
  "type": "local",
  "command": [
    "/opt/quality-coach-mcp/.venv/bin/python",
    "-m",
    "quality_coach_mcp.server"
  ],
  "cwd": "/opt/quality-coach-mcp",
  "enabled": true
}

Herstart OpenCode na wijzigingen:

systemctl restart opencode-serve-4096

Gebruik

Via OpenCode chat

De tools zijn beschikbaar als MCP-tools in elke OpenCode-sessie. Voorbeelden:

  • "Wat is de test coverage van /srv/cerios-clinic?"

  • "Geef een maturity assessment voor project X met scores {'test_planning': 3, 'test_design': 2}"

  • "Genereer een quality report voor gateway met een periode van 30 dagen"

  • "Wat zegt ISO 25010 over Maintainability?"

  • "Check de quality gate pull_request voor Steavy/cerios-clinic"

Directe Python API

from quality_coach_mcp.tools.code_analysis import CodeAnalysisTools
from quality_coach_mcp.config import load_config

config = load_config()
tools = CodeAnalysisTools(config)

import asyncio
result = asyncio.run(tools.analyze_code_quality("/srv/cerios-clinic"))
print(result)

Kennisbank

De server bevat een embedded kennisbank met:

  • ISO 25010 — 8 kwaliteitskenmerken met sub-characteristics

  • TMMi — 5 maturity levels met assessment areas

  • Test Patterns — Herkenbare patronen en anti-patterns

  • Quality Gates — Standaard criteria voor PR, release, sprint

  • Recommendations — Context-gevoelige verbeteradviezen

  • CI/CD Readiness — Maturitychecklist voor CI/CD-processen (102 vragen)

  • AI Readiness — Maturitychecklist voor AI-toepassing (60 vragen, 5 domeinen)

Kennisbank-bestanden staan in src/quality_coach_mcp/knowledge/ als YAML.

Development

Tests draaien

pip install -e ".[dev]"
pytest tests/ -v

Projectstructuur

src/quality_coach_mcp/
├── server.py              # FastMCP server entry point (16 tools)
├── config.py              # Config loading
├── models/types.py        # Pydantic data models
├── tools/
│   ├── code_analysis.py   # Code kwaliteitsanalyse
│   ├── issue_analysis.py  # Issue & defect analyse
│   ├── cicd_tools.py      # CI/CD pipeline tools
│   └── knowledge_tools.py # Kennis & maturity tools
├── adapters/
│   ├── github_adapter.py  # GitHub API wrapper
│   └── cicd_adapter.py    # CI/CD resultaten parser
└── knowledge/
    ├── iso25010.yaml
    ├── tmmi.yaml
    ├── test_patterns.yaml
    ├── quality_gates.yaml
    ├── recommendations.yaml
    ├── cicd_readiness.yaml
    └── ai_readiness.yaml

Licentie

Onderdeel van Cerios TechLab.

Available Tools

16 tools
ai_readiness_scanAi Readiness ScanB

Run an AI Readiness Scan assessment with 60 questions across 5 AI domains.

Inputs: project (required), answers. Returns: object

analyze_code_qualityAnalyze Code QualityB

Scan a repository for code quality metrics (complexity, duplication, smells).

Inputs: repo_path (required), branch. Returns: object

analyze_flaky_testsAnalyze Flaky TestsC

Detect flaky tests from CI run data.

Inputs: github_ref (required), run_ids. Returns: object

analyze_test_coverageAnalyze Test CoverageB

Analyze test coverage by comparing test files against source files.

Inputs: repo_path (required). Returns: object

cicd_readiness_scanCicd Readiness ScanB

Run a CI/CD Readiness Scan assessment with 102 questions across 5 domains.

Inputs: project (required), answers. Returns: object

defect_trend_analysisDefect Trend AnalysisB

Analyze issue trends (volume, resolution time, severity) for a repository.

Inputs: github_ref (required), days. Returns: object

detect_test_patternsDetect Test PatternsC

Detect test patterns and anti-patterns in a repository.

Inputs: repo_path (required). Returns: object

framework_lookupFramework LookupA

Look up information about a quality framework (ISO 25010, TMMi).

Inputs: framework (required), topic. Returns: object

generate_quality_reportGenerate Quality ReportC

Generate a comprehensive Markdown quality report.

Inputs: project (required), period, coverage_pct, maturity_scores, +1 more. Returns: result

maturity_assessmentMaturity AssessmentC

Run a TMMi maturity assessment for a project.

Inputs: project (required), scores. Returns: object

pipeline_healthPipeline HealthA

Analyze CI/CD pipeline health (success rate, duration trends).

Inputs: github_ref (required), days. Returns: object

quality_gate_checkQuality Gate CheckC

Check if quality gates are met for a project.

Inputs: github_ref (required), gate_id, metrics. Returns: object

quality_hotspot_detectionQuality Hotspot DetectionB

Identify files with the highest defect density (hotspots).

Inputs: github_ref (required). Returns: object

quality_recommendationsQuality RecommendationsC

Generate prioritised quality improvement recommendations.

Inputs: maturity_scores (required), context. Returns: object

root_cause_categoriesRoot Cause CategoriesB

Categorise closed bugs by root cause (code, design, requirements, security).

Inputs: github_ref (required). Returns: object

test_result_summaryTest Result SummaryA

Get test results summary for a specific CI workflow run.

Inputs: run_id (required), github_ref (required). Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.3/5.0

    Scored across 16 tools

    Disambiguation4/5

    Each tool targets a distinct aspect of quality (pipeline health, test results, code quality, readiness scans, defects), and the descriptions help differentiate them. There is mild overlap between analysis tools like analyze_code_quality and quality_hotspot_detection, but their scopes are clear enough to avoid major misselection.

    Naming Consistency3/5

    All names are snake_case and descriptive, making the set readable. However, the convention is mixed: some tools use noun phrases (pipeline_health, framework_lookup) while others use verb-led names (analyze_code_quality, detect_test_patterns, generate_quality_report), so the naming pattern is not fully consistent.

    Tool Count4/5

    At 16 tools, the set is slightly over the ideal 3-15 range but still reasonable given the broad quality transformation domain. Each tool earns its place by covering a distinct part of the assessment, analysis, and reporting workflow.

    Completeness4/5

    The tool surface covers assessment, analysis, recommendations, and report generation across code quality, testing, CI/CD, and maturity frameworks. A minor gap is the lack of operational follow-up tools such as tracking remediation progress, but the core workflow is well covered.

    Maintenance

    ActivitySlowing
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers