Skills
Navigate
skills/riffingonsoftware/agent-skills/open-pr/Gen Agent Trust Hub

open-pr

Pass

Audited by Gen Agent Trust Hub on Nov 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes repository content including diffs and untracked files, which could theoretically contain malicious instructions. However, the risk is mitigated by instructions that direct the agent to verify changes and stop if sensitive data or ambiguous scopes are encountered. \n
  • Ingestion points: SKILL.md (repository instructions, branch ancestry, staged/unstaged diffs, untracked files).\n
  • Boundary markers: Absent in the prompt templates.\n
  • Capability inventory: SKILL.md (git commit/push via ship and PR management via gh CLI).\n
  • Sanitization: The agent is instructed to inspect content for sensitive data before proceeding.\n- [COMMAND_EXECUTION]: The skill utilizes several CLI tools to perform its tasks, including gh for GitHub operations and custom tools like deslop, unslop, ship, and babysit-pr for code and prose processing. These appear to be utility tools intended for the stated PR management purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 6, 2026, 12:50 PM