create-github-actions

Pass

Audited by Gen Agent Trust Hub on Mar 17, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses common shell utilities (ls, cat, grep, jq, docker) to analyze project structure and configuration files during the workflow generation process. These commands are executed locally to gather metadata for template customization.
  • [EXTERNAL_DOWNLOADS]: The generated workflow templates reference well-known and widely-trusted GitHub Actions, such as actions/checkout, shivammathur/setup-php, actions/cache, codecov/codecov-action, and aquasecurity/trivy-action. These are industry-standard tools for PHP CI/CD pipelines.
  • [CREDENTIALS_UNSAFE]: Documentation examples in references/examples.md include default placeholder credentials for local development services (e.g., mysql://root:root@127.0.0.1:3306/test and amqp://guest:guest@127.0.0.1:5672). These are standard test-environment defaults and do not represent a security risk.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by ingesting metadata from composer.json to customize workflows.
  • Ingestion points: composer.json file content (referenced in references/examples.md).
  • Boundary markers: Absent.
  • Capability inventory: Filesystem read/write and execution of local analysis tools.
  • Sanitization: None identified; the skill assumes the project configuration follows standard JSON structure.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 17, 2026, 09:58 AM