k2f-survey-form
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user data to generate survey blueprints, creating a surface for indirect prompt injection.
- Ingestion points: User inputs describing the survey purpose, target audience, measurement constructs, and question details (Step 1.1, 1.2).
- Boundary markers: The skill instructs the agent to use specific placeholders like
OPEN QUESTIONorNEEDS USER INPUTfor missing or sensitive data instead of fabricating it, which serves as a basic safety delimiter. - Capability inventory: The skill interacts with the
k2fengine to process content and design intent, eventually exporting files to formats like PDF or DOCX (Step 2). - Sanitization: The workflow mandates a 'Truth gate' and 'Question design pass' to ensure organization names and claims trace back to user input, and it emphasizes neutral stems to mitigate leading prompts.
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install a core dependency using the command
npx skills add suzheng/k2f --skill k2f. This resource originates from the same vendor (suzheng) and is considered a functional requirement.
Audit Metadata