ios-dev-guidelines
Pass
Audited by Gen Agent Trust Hub on May 3, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill establishes a surface for indirect prompt injection by instructing the agent to process and refactor external Swift source code files.
- Ingestion points: The agent processes .swift files and related documentation triggered by specific architectural keywords.
- Boundary markers: The instructions do not define any delimiters or explicit commands to ignore potential instructions embedded within code comments or strings.
- Capability inventory: The agent has the ability to search codebases using ripgrep (rg), modify/refactor code, and delete files (rm).
- Sanitization: There are no specified sanitization or validation mechanisms for the content of the processed files.
- [COMMAND_EXECUTION]: The skill documentation acknowledges and constrains the agent's ability to execute shell commands (e.g., rg, rm, git). It specifically provides safety guidelines to prevent historical errors like recursive wildcard deletion or autonomous commits without user approval.
Audit Metadata