ios-debugger-agent
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as build error logs and application UI accessibility labels via the
mcp__XcodeBuildMCP__describe_uitool. This represents a potential surface for indirect prompt injection where malicious content in logs or UI elements could influence agent behavior. However, this ingestion is necessary for the skill's primary purpose as a debugger, and no active exploits or deceptive instructions were identified. - Ingestion points: Build failure error output and simulator UI hierarchy descriptions.
- Boundary markers: Not explicitly defined in instructions.
- Capability inventory: Ability to execute builds, launch applications, capture screenshots, and perform UI interactions (tap/type).
- Sanitization: No specific sanitization or filtering of external data strings is observed.
- [DATA_EXPOSURE]: The skill accesses local file system paths (projectPath/workspacePath) and captures simulator logs and screenshots. These actions are within the expected scope of an iOS debugger agent and do not constitute unauthorized data exposure.
Audit Metadata