orient
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell commands including 'gh issue view', 'ls', and 'tail' to retrieve issue metadata, search for Architecture Decision Records (ADRs), and read recent memory logs from the workspace.
- [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection by ingesting and acting upon untrusted data from external GitHub issues.
- Ingestion points: Output of 'gh issue view' containing issue titles and bodies (SKILL.md).
- Boundary markers: Not present; issue content is integrated directly into the orientation process without delimiters or instructions to ignore embedded commands.
- Capability inventory: File system listing (ls), file reading (tail), and GitHub CLI interactions (gh).
- Sanitization: No sanitization or validation of the fetched issue content is performed, allowing potentially malicious descriptions to influence routing and task classification logic.
Audit Metadata