factory-ship
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill defines a routine development workflow for continuous integration and delivery. It references documentation from the author's official GitHub repository and uses local configuration files to manage tasks, which is standard practice for this type of agent skill.
- [COMMAND_EXECUTION]: The verification step involves executing local tools such as formatters, test suites, and release checks based on the project's specific configuration.
- [INDIRECT_PROMPT_INJECTION]: The skill reads external data from
.agent-factory/config.yamland repository instructions, creating a surface where content could influence agent behavior. - Ingestion points:
.agent-factory/config.yaml,skill_prompts.factory-shipconfiguration, and repository-level instructions. - Boundary markers: The instructions direct the agent to specific configuration entries and provide guidance on authorization requirements before proceeding.
- Capability inventory: Running local verification commands, managing PR lifecycle (title, body, labels), and performing branch merges.
- Sanitization: The skill relies on configured delivery policies and explicit authorization gates rather than automated sanitization of ingested text.
Audit Metadata