git-branch
Pass
Audited by Gen Agent Trust Hub on Nov 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands such as
gh issue viewandgit switchusing variables derived from user input. Whilereferences/issue-id-examples.mdprovides patterns for parsing, the lack of explicit shell-safe sanitization for the<issueID>and<branch-name>variables poses a risk of command injection if the agent interpolates unvalidated strings into the shell environment.- [INDIRECT_PROMPT_INJECTION]: The skill incorporates untrusted data from external sources into the agent's context. \n - Ingestion points: Fetches issue title and body content using the
gh issue viewcommand inSKILL.md. \n - Boundary markers: None; the skill does not use delimiters or instructions to ignore embedded commands within the fetched issue text. \n
- Capability inventory: The skill can perform repository operations including
git switchandgit check-ref-formatas defined inSKILL.md. \n - Sanitization: The skill provides instructions to format the output branch name (lowercase, removing punctuation, hyphen-separated), which acts as a filter for the resulting command but does not protect the agent context from the initial ingestion.
Audit Metadata