work-define
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides the agent with shell commands to resolve environment variables and determine the active git branch. Specifically, it uses
git branch --show-currentto prevent mixing documentation updates with implementación branches and a shell script to resolve theCLAUDE_PLUGIN_ROOTpath for shared references. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted functional requirements from user descriptions and Software Requirement Specification (SRS) documents to generate User Stories.
- Ingestion points: Direct user input for functional needs and the contents of
README.mdfiles located inrequirements/SRS-XXX-[nombre-corto]/. - Boundary markers: None identified; the skill relies on the LLM to follow the provided templates without explicit delimiters for the ingested untrusted content.
- Capability inventory: The skill performs file system writes to create and update documentation and utilizes subagent chaining to invoke related skills like
/design-define,/test-define, and/work-plan. - Sanitization: There is no specific instruction to sanitize or filter the ingested data for potential prompt injection or adversarial content.
Audit Metadata