feedback
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user input, specifically verbatim phrases and descriptions of user experiences, to file GitHub issues.
- Ingestion points: User feedback strings and verbatim quotes are captured and processed as described in the
SKILL.mdinstructions. - Boundary markers: The skill uses shell heredocs with quoted delimiters (
'EOF') for the issue body and double-quotes for command-line arguments to separate data from instructions. - Capability inventory: The skill uses the GitHub CLI (
gh) to perform repository views, list issues, create labels, and file new issues or comments. - Sanitization: Employs quoted heredocs and argument quoting to prevent the execution of shell meta-characters contained within user input.
- [COMMAND_EXECUTION]: Executes shell commands via the GitHub CLI (
gh) to manage issues and labels. This behavior is consistent with the skill's documented purpose of providing a feedback loop for developers.
Audit Metadata