new-compiler-feature
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- Command Execution via CLI Tools: The skill relies on the
gh(GitHub) CLI to perform administrative actions on repositories, such as creating labels, issues, and comments. This involves executing shell commands where parameters are dynamically populated from external data. - Indirect Prompt Injection Surface: The skill ingests untrusted data from GitHub Pull Request bodies and issue titles to determine feature names and URLs. This represents an attack surface where maliciously crafted external content could attempt to influence the agent's behavior.
- Ingestion points: Data is gathered using
gh pr viewandgh issue viewin Phase 1. - Boundary markers: The skill does not use explicit delimiters to isolate untrusted data when interpolating it into command strings.
- Capability inventory: The skill possesses write capabilities, including
gh label create,gh issue create,gh pr comment, and file updates viagh api. - Sanitization: While the skill asks the user to confirm or edit feature names, it does not perform automated sanitization of the input strings before interpolating them into shell commands.
- Dynamic Command Assembly: The skill constructs PowerShell scripts and API calls at runtime using metadata gathered from the implementation PR. If these variables contain shell-active characters (e.g., backticks or semicolons), they could lead to unintended command execution.
Audit Metadata