create-watchos-version
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (scripts/analyze_project.py) to analyze the directory structure and contents of a target project. This script performs file system traversal and reads the contents of configuration and source files.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the user's local file system (Swift source code, project configuration) to generate implementation plans. This creates an attack surface where a malicious project could contain instructions intended to influence the agent's behavior.
- Ingestion points: The script scripts/analyze_project.py reads content from *.swift, *.xcodeproj, Package.swift, and Info.plist files within the project root.
- Boundary markers: The skill instructions do not explicitly mandate the use of delimiters or warnings to disregard instructions found within the ingested source code during plan generation.
- Capability inventory: The agent has the capability to read local files, execute analysis scripts, and generate structured implementation plans based on that data.
- Sanitization: The Python analysis script uses regex to identify patterns but does not sanitize or filter the raw data extracted from the project files before it is processed by the agent.
Audit Metadata