Skills
Navigate
skills/shibayu36/agent-skills/my-create-draft-pull-request/Gen Agent Trust Hub

my-create-draft-pull-request

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from existing Pull Request descriptions and repository templates to inform the creation of new Pull Requests. 1. Ingestion points: SKILL.md instructs the agent to read .github/pull_request_template.md and the output of ghro pr view <number> (which retrieves PR bodies). 2. Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard potential commands embedded in the ingested PR content. 3. Capability inventory: The skill uses git push, ghro (GitHub CLI wrapper), and open. 4. Sanitization: The skill lacks sanitization for the content retrieved from external PRs before it is processed by the main agent or the review SubAgent.
  • [COMMAND_EXECUTION]: The skill utilizes non-standard command-line utilities gitro and ghro. These tools appear to be wrappers for git and gh (GitHub CLI). While likely functional for repository interaction in the user's environment, they are not standard system commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 12:59 AM