github-project-automation

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes several shell scripts (generate-codeowners.sh, setup-github-project.sh, sync-templates.sh, validate-workflows.sh) that automate repository configuration. These scripts perform local file operations (copying, directory creation) and execute standard version control commands using git and the GitHub CLI (gh).
  • [EXTERNAL_DOWNLOADS]: The workflow templates reference official and well-known GitHub Actions to perform tasks like code checkout, environment setup, and security scanning.
  • Fetches official actions from the GitHub-managed actions and github organizations (e.g., actions/checkout, actions/setup-node, github/codeql-action).
  • Integrates with well-known third-party services such as codecov/codecov-action for code coverage reporting.
  • All actions in the provided templates are pinned to cryptographic SHAs, following security best practices to prevent supply chain attacks.
  • [PRIVILEGE_ESCALATION]: The SKILL.md file provides instructions for installing the GitHub CLI using sudo apt install gh. This is a standard administrative procedure for package installation on Linux systems and is necessary for the skill's functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill's automation scripts ingest data from the repository's git history and environment.
  • Ingestion points: pr-checks.yml reads branch names and commit messages from the github.head_ref and github.event contexts; generate-codeowners.sh parses git logs.
  • Boundary markers: None explicitly used for prompt delimiters.
  • Capability inventory: The skill has the capability to write to the local file system and execute shell commands via the provided scripts and workflows.
  • Sanitization: The scripts employ regular expressions to validate branch naming conventions and commit message formats, reducing the risk of processing malicious metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 01:42 PM
Security Audit — agent-trust-hub — github-project-automation