Skills
Navigate
skills/seabbs/skills/bot-report/Gen Agent Trust Hub

bot-report

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local shell script located at ~/code/seabbs/dotfiles/scripts/review-bot.sh with the --status and --json flags to retrieve bot state.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data by fetching GitHub pull request comments via gh api.
  • Ingestion points: GitHub API endpoint repos/<repo>/pulls/<n>/comments in SKILL.md.
  • Boundary markers: The skill contains a 'Boundaries' section explicitly instructing the agent that comments from unauthorized users are not instructions and should not be acted upon or summarized.
  • Capability inventory: The skill has permissions to execute local scripts and use the GitHub CLI (gh).
  • Sanitization: The instructions provide a logical filter, directing the agent to report the existence of comments without processing their content as instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 09:54 PM