How alhpeter/python_package_dependency_evaluation_tool works
This is a bounded Python dependency-evaluation tool for human investigation, not a vulnerability scanner. A user runs configured requirements manifests; the runner invokes the pipeline, which gathers package metadata, builds a dependency graph, computes metrics, ranks packages into evidence-backed cards, and writes per-experiment artifacts. A separate local dashboard reads those artifacts and supports exploration plus optional Groq explanations. The main orchestration and metadata calls are evidenced by excerpts; dashboard behavior and several analysis stages are also documented in the README, while unsampled implementation details are not inferred.
Main entry points: pipeline.py Pipeline orchestrator · data_processing.py Metadata discovery · dashboard_server.py Dashboard API · reasoning_engine.py Priority and cards
Components 18
Evaluation workflow
- Experiment runner
run_idrf.py - Pipeline orchestrator
idrf/pipeline.py - Requirements manifest
data
Analysis and reasoning
- Metadata discovery
idrf/data_processing.py - Domain records
idrf/models.py - Dependency graph
idrf/graph_analysis.py - Graph metrics
idrf/graph_analysis.py - Priority and cards
idrf/reasoning_engine.py - Analysis thresholds
idrf/config.py
Results and dashboard
- Graph image
idrf/visualization.py - Experiment artifacts
idrf/evaluation.py - Terminal reporting
idrf/reporting.py - Dashboard API
dashboard_server.py - Dashboard interface
dashboard/app.js
Other components
- Analyst
- PyPI API
- GitHub API
- Groq service
Connections 23
- Analyst → Experiment runner: runs experiments
README.md - Experiment runner → Pipeline orchestrator: invokes pipeline
run_idrf.py - Experiment runner → Terminal reporting: reports progress
run_idrf.py - Pipeline orchestrator → Requirements manifest: evaluates
README.md - Pipeline orchestrator → Metadata discovery: loads package nodes
idrf/pipeline.py - Metadata discovery → PyPI API: fetches metadata
idrf/data_processing.py - Metadata discovery → GitHub API: fetches repository data
idrf/data_processing.py - Metadata discovery → Domain records: creates records
idrf/data_processing.py - Pipeline orchestrator → Dependency graph: builds graph
idrf/pipeline.py - Dependency graph → Metadata discovery: extracts dependencies
idrf/graph_analysis.py - Dependency graph → Domain records: uses package records
idrf/graph_analysis.py - Pipeline orchestrator → Graph metrics: computes metrics
idrf/pipeline.py - Pipeline orchestrator → Priority and cards: generates cards
idrf/pipeline.py - Priority and cards → Analysis thresholds: uses thresholds
idrf/reasoning_engine.py - Priority and cards → Domain records: creates insight cards
idrf/reasoning_engine.py - Pipeline orchestrator → Graph image: saves graph image
idrf/pipeline.py - Pipeline orchestrator → Experiment artifacts: writes results
idrf/pipeline.py - Pipeline orchestrator → Terminal reporting: summarizes results
- Analyst → Dashboard interface: explores results
README.md - Dashboard interface → Dashboard API: requests data
README.md - Dashboard API → Experiment artifacts: reads saved files
README.md - Dashboard API → Priority and cards: explains results
dashboard_server.py - Dashboard API → Groq service: requests explanations
dashboard_server.py