How kesavanaswin/cspm works
CSPM analyzes container configuration, vulnerability findings, and runtime telemetry, correlates them into risk assessments, then simulates admission decisions and alerts. The primary documented workflow is the modular CLI/Streamlit pipeline; the repository also contains a separate Flask-oriented scan workflow with config auditing, scanner/monitor integrations, and its own risk-scoring modules. The README documents the modular stages; source excerpts show key orchestration and implementation details. Other relationships are limited to imports recorded in the SOURCE INDEX.
Main entry points: app.py Flask API · main.py CLI runner · gui.py Streamlit dashboard · analysis_engine.py Policy and anomaly analysis
Components 20
User Interfaces
- CLI runner
main.py - Streamlit dashboard
gui.py - Flask API
app.py
CSPM Pipeline
- Input ingestion
engines/ingestion_engine.py - Policy and anomaly analysis
engines/analysis_engine.py - Unified risk scoring
engines/risk_engine.py - Admission and alerts
engines/mitigation_engine.py
Flask Scan Workflow
- Manifest config audit
config_audit.py - Image vulnerability scan
vuln_scan.py - Runtime monitoring
runtime_monitor.py - Legacy risk correlation
risk_engine.py - Scan records
models.py
Shared Infrastructure
- JSON persistence
storage/json_store.py - Pipeline data models
utils/models.py - Central logging
utils/logger.py
Policy Inputs
- Policy rules
config/policies.yaml - Rego policies
policies/container_security.rego - Scanner and risk config
config.py
Other components
- Operator
- Docker/Kubernetes
Connections 31
- Operator → CLI runner: runs analysis
README.md - Operator → Streamlit dashboard: views dashboard
README.md - Operator → Flask API: submits scan
app.py - Docker/Kubernetes → Input ingestion: provides state
- CLI runner → Input ingestion: loads inputs
main.py - CLI runner → Policy and anomaly analysis: evaluates findings
README.md - CLI runner → Unified risk scoring: assesses risk
main.py - CLI runner → Admission and alerts: processes assessment
main.py - Streamlit dashboard → Input ingestion: loads inputs
gui.py - Streamlit dashboard → Policy and anomaly analysis: evaluates findings
gui.py - Streamlit dashboard → Unified risk scoring: assesses risk
gui.py - Streamlit dashboard → Admission and alerts: processes assessment
gui.py - Input ingestion → JSON persistence: loads JSON
engines/ingestion_engine.py - Input ingestion → Pipeline data models: constructs records
engines/ingestion_engine.py - Policy and anomaly analysis → Pipeline data models: evaluates records
engines/analysis_engine.py - Policy and anomaly analysis → Policy rules: loads rules
README.md - Policy and anomaly analysis → Central logging: logs findings
engines/analysis_engine.py - Unified risk scoring → Pipeline data models: scores assessments
engines/risk_engine.py - Admission and alerts → JSON persistence: persists alerts
engines/mitigation_engine.py - Admission and alerts → Pipeline data models: builds alerts
engines/mitigation_engine.py - Admission and alerts → Central logging: logs actions
engines/mitigation_engine.py - Flask API → Manifest config audit: audits manifest
app.py - Flask API → Image vulnerability scan: scans image
app.py - Flask API → Runtime monitoring: checks events
app.py - Flask API → Legacy risk correlation: computes risk
app.py - Flask API → Scan records: records findings
app.py - Manifest config audit → Scanner and risk config: reads settings
config_audit.py - Image vulnerability scan → Scanner and risk config: reads settings
vuln_scan.py - Runtime monitoring → Scanner and risk config: reads settings
runtime_monitor.py - Legacy risk correlation → Scanner and risk config: uses weights
risk_engine.py - Manifest config audit → Rego policies: evaluates policies