comment-writer

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, such as GitHub PRs, issues, and Slack/Discord messages, to generate collaboration comments. This creates a surface where malicious instructions embedded in those sources could influence the agent's behavior.
  • Ingestion points: The skill triggers on PR feedback, issue replies, and chat messages. Specifically, the gh pr view command in SKILL.md fetches title, body, and code changes from GitHub.
  • Boundary markers: There are no explicit instructions or delimiters telling the agent to ignore or isolate instructions that might be contained within the external content.
  • Capability inventory: The skill uses the gh (GitHub CLI) tool to execute commands and retrieve remote data.
  • Sanitization: No sanitization, filtering, or validation steps are defined for the incoming text before the agent applies the Voice Rules.
  • [COMMAND_EXECUTION]: The skill utilizes the GitHub CLI to inspect pull requests.
  • Evidence: gh pr view <PR_NUMBER> --json title,body,additions,deletions,changedFiles in SKILL.md allows the agent to execute shell commands to interact with GitHub repositories.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 03:31 PM
Security Audit — agent-trust-hub — comment-writer