The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes instructions to execute a provided Python script scripts/validate_requirements_contract.py for validating manifest files. The script is a local utility that performs static analysis on JSON data and does not perform network operations or unsafe system calls.
[INDIRECT_PROMPT_INJECTION]: The workflow involves processing external evidence and product requirements. This ingestion of untrusted data represents an indirect prompt injection surface; however, the lack of dangerous capabilities (network, write access) and the use of explicit validation steps mitigate this risk.
Ingestion points: Elicitation evidence and product goals (SKILL.md)
Boundary markers: Absent
Capability inventory: Local script execution (scripts/validate_requirements_contract.py)
Sanitization: Absent

requirements-definition