Skip to main content
Filter commands help solve the problem of visibility for OSS dependencies in applications. To support various requirements, we adopt the generic Common Expression Language (CEL) for flexible filtering.

​
How Filtering Works

Filter expressions work on packages (dependencies) and evaluate to a boolean result. Packages are included in results when expressions evaluate to true.

​
Quick Examples

​
License Filtering

List only packages that use the MIT license: 
vet scan -D /path/to/repo \
    --report-summary=false \
    --filter 'licenses.exists(p, p == "MIT")'

​
Popularity Analysis

Find dependencies that appear unpopular: 
vet scan --lockfiles /path/to/pom.xml --report-summary=false \
    --filter='projects.exists(x, x.stars < 10)'

​
Vulnerability Detection

Find dependencies with critical vulnerabilities: 
vet scan --lockfiles /path/to/pom.xml --report-summary=false \
    --filter='vulns.critical.exists(x, true)'

​
Filter Input Structure

Filter expressions receive the following input data:
VariableContent
_The root variable, holding other variables
vulnsMap of vulnerabilities organized by severity
scorecardOpenSSF Scorecard metrics and scores
projectsList of source projects associated with the package
licensesList of licenses in SPDX license code format
pkgPackage information (ecosystem, name, version)
Refer to the filter input specification for the detailed structure of input messages.

​
Common Filter Expressions

​
Security-Focused Filters

# Critical vulnerabilities
vulns.critical.exists(x, true)

# Any high or critical vulnerabilities
vulns.critical.size() > 0 || vulns.high.size() > 0

# More than 5 medium vulnerabilities
vulns.medium.size() > 5

# No known vulnerabilities
vulns.critical.size() == 0 && vulns.high.size() == 0 && 
vulns.medium.size() == 0 && vulns.low.size() == 0

​
Project Quality Filters

# Low popularity packages
projects.exists(x, x.stars < 100)

# High maintenance burden (many open issues)
projects.exists(x, x.issues > 500)

# Well-maintained projects
projects.exists(x, x.stars > 1000 && x.forks > 100)

# Community engagement ratio
projects.exists(x, x.forks > 0 && (x.stars / x.forks) > 5)

​
Advanced Combination Filters

# High-risk packages
vulns.critical.size() > 0 && 
scorecard.scores.Maintained == 0 && 
projects.exists(x, x.stars < 50)

# Compliance violations
!licenses.exists(p, p in ["MIT", "Apache-2.0", "BSD-3-Clause"]) ||
vulns.high.size() > 0 ||
scorecard.Score < 3

# Production-ready packages
vulns.critical.size() == 0 &&
vulns.high.size() == 0 &&
scorecard.scores.Maintained > 5 &&
projects.exists(x, x.stars > 500) &&
licenses.exists(p, p in ["MIT", "Apache-2.0"])

​
Filter Input JSON Example

Understanding the structure helps create effective filters: 
{
  "pkg": {
    "ecosystem": "npm",
    "name": "lodash.camelcase",
    "version": "4.3.0"
  },
  "vulns": {
    "all": [],
    "critical": [],
    "high": [],
    "medium": [],
    "low": []
  },
  "scorecard": {
    "scores": {
      "Binary-Artifacts": 10,
      "Branch-Protection": 0,
      "CII-Best-Practices": 0,
      "Code-Review": 8,
      "Dangerous-Workflow": 10,
      "Dependency-Update-Tool": 0,
      "Fuzzing": 0,
      "License": 10,
      "Maintained": 0,
      "Packaging": -1,
      "Pinned-Dependencies": 9,
      "SAST": 0,
      "Security-Policy": 10,
      "Signed-Releases": -1,
      "Token-Permissions": 0,
      "Vulnerabilities": 10
    }
  },
  "projects": [
    {
      "name": "lodash/lodash",
      "type": "GITHUB",
      "stars": 55518,
      "forks": 6787,
      "issues": 464
    }
  ],
  "licenses": ["MIT"]
}

​
Practical Use Cases

​
Security Auditing

# Generate security report
vet scan -D . \
    --filter 'vulns.critical.size() > 0 || vulns.high.size() > 0' \
    --report-json security-issues.json

# Find packages needing immediate attention
vet scan -D . \
    --filter 'vulns.critical.size() > 0 && scorecard.scores.Maintained == 0' \
    --report-markdown urgent-issues.md

​
License Compliance

# Check for license violations
vet scan -D . \
    --filter '!licenses.exists(p, p in ["MIT", "Apache-2.0", "BSD-3-Clause", "ISC"])' \
    --report-summary

# Find packages without license information
vet scan -D . \
    --filter 'licenses.size() == 0' \
    --report-json unlicensed-packages.json

​
Quality Assessment

# Find low-quality dependencies
vet scan -D . \
    --filter 'projects.exists(x, x.stars < 100) && scorecard.scores.Maintained < 5' \
    --report-summary

# Identify potential supply chain risks
vet scan -D . \
    --filter 'scorecard.scores.Supply_Chain < 5 || scorecard.scores.Token_Permissions < 5' \
    --report-json supply-chain-risks.json

​
Performance Tips

  • Place most selective conditions first
  • Use && instead of nested conditions when possible
  • Avoid redundant checks in complex expressions
  • Use size() for counting array elements
  • Use exists() for checking array membership
  • Use direct comparison for simple values
  • Combine multiple similar checks into single expressions
  • Use the query workflow for repeated filtering operations
  • Cache results using JSON dumps for multiple analyses

​
CEL Language Quick Reference

​
Common Functions

  • size() - Get length of arrays or maps
  • exists(var, condition) - Check if condition is true for any element
  • in - Check membership in arrays or maps
  • contains() - String contains check
  • startsWith() / endsWith() - String prefix/suffix checks

​
Operators

  • ==, != - Equality comparison
  • <, <=, >, >= - Numeric comparison
  • &&, || - Logical AND, OR
  • ! - Logical NOT
  • +, -, *, / - Arithmetic operations

​
Data Types

  • true, false - Booleans
  • "string" - Strings (double quotes)
  • 123, 45.67 - Numbers
  • ["a", "b"] - Arrays
  • {"key": "value"} - Maps

CEL Documentation

Complete Common Expression Language documentation

OpenSSF Scorecard

Learn about available scorecard checks and metrics

Build Your Own Queries

Use filtering with the query workflow for better performance

Policy as Code

Implement comprehensive policies using filter expressions