An AI-powered agent for validating nf-core pipeline compliance. This tool analyzes Nextflow pipelines against the official nf-core guidelines and provides detailed recommendations for improving compliance.
- Documentation-Based Validation: Uses the latest nf-core guidelines directly from the official documentation
- Comprehensive Analysis: Checks modules, workflows, and subworkflows against all relevant requirements
- Detailed Reports: Generates comprehensive reports with specific recommendations for each component
- AI-Powered: Leverages large language models to understand complex requirements and provide context-aware suggestions
- Interactive Chat: Query the nf-core documentation directly with natural language questions
- Categorized Results: View validation results and chat responses organized by documentation category
- Python 3.8 or higher
- OpenAI API key
# Clone the repository
git clone https://github.com/arunbodd/nf-core_guidelines_validator
cd nf-core_guidelines_validator
# Install the package
pip install -e .Before validating pipelines, you need to harvest the nf-core documentation:
# Set your OpenAI API key
export OPENAI_API_KEY="your-api-key"
# Harvest documentation
nfcore-validator harvestThis will create a vector store of nf-core guidelines in the current directory.
# Validate a pipeline
nfcore-validator validate /path/to/your/pipeline --format markdownThis will:
- Analyze all components in the pipeline
- Check them against relevant nf-core guidelines
- Generate a detailed compliance report
You can also directly query the nf-core documentation using natural language:
# Start the chat interface
nfcore-validator chat
# Show sources for answers
nfcore-validator chat --show-sources
# Increase context size for more comprehensive answers
nfcore-validator chat --context-size 15The tool generates two types of reports:
- JSON Report: Contains all validation details in a structured format
- Markdown Report: Human-readable summary with component details and recommendations
The enhanced markdown report includes:
- Component Type Breakdown: Shows compliance by component type
- Detailed Violations: Descriptions and example fixes for each violation type
- Organized Components: Grouped by type with failed requirements highlighted
- Actionable Recommendations: Prioritized list of improvements with affected components
Example markdown report:
# nf-core Pipeline Compliance Report
**Pipeline:** `rnaseq`
**Path:** `/path/to/pipeline`
**Date:** 2025-04-28 00:11:25
## Summary
- **Components Analyzed:** 25
- **Requirements Checked:** 143
- **Passed Requirements:** 128
- **Failed Requirements:** 15
- **Compliance Score:** 89.5%
## Component Type Breakdown
| Component Type | Count | Avg. Compliance |
|---------------|-------|----------------|
| module | 15 | 92.33% |
| workflow | 5 | 85.60% |
...nfcore-validator harvest --output /path/to/vectorstore
nfcore-validator validate /path/to/pipeline --vectorstore /path/to/vectorstoreThe validator automatically handles OpenAI API rate limits by:
- Processing components gradually
- Waiting and retrying when rate limits are hit
- Adding small delays between API calls
You can also reduce parallelism to further avoid rate limits:
python -m nfcore_validator.cli.main validate /path/to/pipeline --max-workers 2The chat interface categorizes information by documentation section:
nfcore-validator chat --show-sourcesThis will display sources grouped by categories like:
- Module Guidelines
- Subworkflow Guidelines
- Pipeline Structure
- Test Data Guidelines
- Documentation Harvesting: The tool extracts all guidelines from the official nf-core documentation website
- Vector Embedding: Guidelines are converted to vector embeddings for semantic search
- Component Analysis: For each pipeline component, the tool retrieves relevant guidelines
- AI Validation: An LLM analyzes the component against the guidelines and generates recommendations
- Report Generation: Results are compiled into comprehensive reports
MIT
Contributions are welcome! Please feel free to submit a Pull Request.