skyrim-ship-gate
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied Skyrim mod archives and release trees through various validation gates. \n
- Ingestion points: Mod release archives (zip files) and project directories processed by the validation gates in SKILL.md and scripts/validate_release_tree.py. \n
- Boundary markers: The instructions do not define specific delimiters for separating mod content from agent instructions during manual inspection or audit steps. \n
- Capability inventory: The skill uses file system access, executes local Python scripts, and calls project-specific CLI tools such as forge. \n
- Sanitization: The validate_release_tree.py script filters for specific junk filenames and build artifacts but does not perform content-level sanitization of external text files (e.g., READMEs, configuration files) that the agent may process.
- [SAFE]: The included Python script scripts/validate_release_tree.py implements secure archive handling, including a safe_extract function that correctly validates zip members against path traversal attacks by ensuring extracted paths remain within the destination root.
- [SAFE]: The validation logic for Skyrim-specific structures (plugins, frameworks, assets) follows established community modding practices and does not perform unauthorized network operations, request elevated privileges, or include hidden commands.
Audit Metadata