file-naming-convention
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to perform file renames using
git mvand execute validation steps like type checking, linting, or running tests. These operations are standard for development workflows and are restricted to the skill's primary purpose of repository standardization. - [DATA_EXFILTRATION]: There are no network operations or attempts to access sensitive system files or credentials. All actions are performed locally on the repository structure.
- [PROMPT_INJECTION]: The instructions do not contain any patterns aimed at overriding agent behavior, bypassing safety guidelines, or extracting system prompts.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes repository file names (external data), the logic is constrained to a strict whitelist of allowed suffixes and naming shapes, significantly limiting the surface for malicious instruction injection via filenames.
Audit Metadata