release-skills

Warn

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill generates and executes git and gh commands (e.g., git commit -m "...", gh pr view <number>) where the parameters are extracted from commit subjects and PR metadata. If a commit message or PR title contains shell metacharacters, it could lead to command injection during the automated workflow.
  • [PROMPT_INJECTION]: The skill ingests untrusted data from the repository's commit logs and GitHub's PR API to determine version bumps and generate multi-language changelogs. An attacker who contributes a commit or PR could embed malicious instructions (Indirect Prompt Injection) to influence the agent's release decisions or the content of the generated release documentation.
  • [DATA_EXFILTRATION]: While the skill primarily pushes to 'origin', it reads sensitive repository configuration and history. If the git remote is maliciously modified or if command injection occurs, there is a potential path for data exposure.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 14, 2026, 06:32 AM