ios-slim-bindings
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- Indirect Prompt Injection Surface: The skill facilitates the ingestion of data from native libraries into the agent's context. 1. Ingestion points: Data is received via wrapper methods in DotnetMyBinding.swift. 2. Boundary markers: The templates do not define delimiters for external data. 3. Capability inventory: The skill utilizes project build tools and native library interop. 4. Sanitization: The boilerplate does not include validation for external inputs. This is a standard pattern for development tools where the security of external components is a consideration for the implementer.
- Command Execution: The implementation guide provides shell scripts and instructions for using development tools such as XcodeGen and CocoaPods. These patterns are used for the skill's primary purpose of configuring and building iOS binding projects and utilize established development utilities.
Audit Metadata