portfolio-setup
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local project-init.sh script found in the plugin's directory to create the project's folder structure. It uses a project slug derived from user input, which the agent is instructed to format as a kebab-case string for safety.
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it processes untrusted data from external sources.
- Ingestion points: Company information is extracted from user-provided URLs and uploaded files in the workspace.
- Boundary markers: There are no explicit delimiters defined to encapsulate the extracted data and prevent instructions within that data from being executed.
- Capability inventory: The skill can run bash scripts and write JSON files.
- Sanitization: The agent is tasked with normalizing the company name into a kebab-case slug, which provides basic protection against command injection in script arguments.
Audit Metadata