cawplan-testpoint-generate

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests requirement data (five fields) from the CawPlan API and processes user input to generate test points. While it includes strict instructions (Red Line 0) to avoid fabricating details, it lacks explicit boundary markers or delimiters to isolate this external data from the skill's instructions, creating a surface for potential indirect prompt injection if the requirement text contains adversarial instructions.
  • Ingestion points: Requirement data is fetched via cawplan api GET as described in SKILL.md §2.
  • Boundary markers: No explicit delimiters (e.g., XML tags) are used to wrap the untrusted requirement content within the prompt context.
  • Capability inventory: The skill utilizes cawplan qa-insights testpoints archive to write data to the platform and cawplan api GET to read requirement data.
  • Sanitization: No specific sanitization or filtering of the requirement text is mentioned in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:18 AM
Security Audit — agent-trust-hub — cawplan-testpoint-generate