skills/crowdin/skills/github-action/Gen Agent Trust Hub

github-action

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructs users to manage credentials via GitHub Actions Secrets (CROWDIN_PROJECT_ID, CROWDIN_PERSONAL_TOKEN) and environment variables, which is the industry standard for secure CI/CD workflows.
  • [EXTERNAL_DOWNLOADS]: The skill references the official crowdin/github-action repository and its associated documentation on GitHub. These are trusted, well-known sources belonging to the software's author.
  • [COMMAND_EXECUTION]: While the skill describes using a command input for the Crowdin CLI, these are bounded to legitimate translation tasks (e.g., status, pre-translate) and are executed within a Docker container specifically designed for these tools.
  • [INDIRECT_PROMPT_INJECTION]: The skill correctly warns that git add . in a workflow can commit unintended files if the job is not properly isolated. It provides remediation guidance (using separate jobs or specific checkout strategies) to prevent untrusted build artifacts from entering the repository's translation branch.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 09:10 AM
Security Audit — agent-trust-hub — github-action