requirements-definition

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted evidence and product goals to generate requirement baselines. Ingestion points: Elicitation evidence and product goals referenced in SKILL.md. Boundary markers: Absent; the instructions do not delimit evidence data or warn the agent to ignore embedded instructions. Capability inventory: Execution of a local Python validator script (scripts/validate_requirements_contract.py) and file-writing operations. Sanitization: None detected.
  • [COMMAND_EXECUTION]: The skill involves the execution of a local Python script scripts/validate_requirements_contract.py for manifest validation. This script is benign, performing standard JSON parsing and schema-like validation without any network access or dynamic code evaluation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 08:38 PM