ark-issues
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (SAFE): The skill provides templates for using the
ghCLI tool to manage issues in themckinsey/agents-at-scale-arkrepository. All actions are consistent with the stated purpose of the skill.\n- [PROMPT_INJECTION] (LOW): The skill exhibits an indirect prompt injection surface by ingesting and processing untrusted data from GitHub issue descriptions and comments.\n - Ingestion points: Data is retrieved from GitHub using
gh search issues,gh issue list, andgh issue view.\n - Boundary markers: Absent; no delimiters or warnings are used to differentiate untrusted issue content from agent instructions.\n
- Capability inventory: The skill allows for writing operations including
gh issue create,gh issue comment, andgh issue edit.\n - Sanitization: Absent; the skill does not specify any sanitization or validation of the content retrieved from external GitHub issues.
Audit Metadata