backup
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
Bashtool to check for system dependencies (gh,fswatch) and execute backup operations. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided inputs such as GitHub repository names and sync intervals which are then interpolated into commands. Ingestion points: User responses to
AskUserQuestionand command arguments-rand-i. Boundary markers: None present in the instructions. Capability inventory:Bash,Write, andGlobtools are accessible. Sanitization: No input validation or escaping logic is specified for the user-provided parameters. - [DYNAMIC_EXECUTION]: The skill includes 'Self-Evolving' instructions that direct the agent to use the
Writetool to modify theSKILL.mdfile itself to correct instructions or parameters. This constitutes a self-modifying code pattern where the agent updates its own operational logic at runtime.
Audit Metadata