cloudflare
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill demonstrates how to build applications that ingest and process untrusted data from HTTP requests and AI-driven workflows.
- Ingestion points: Data enters the system through
fetch(request),request.json(), andrequest.urlparameters within Worker and Pages function handlers (SKILL.md). - Boundary markers: The provided implementation examples do not use explicit delimiters or boundary markers to isolate untrusted user data from operational instructions.
- Capability inventory: The skill provides examples for powerful capabilities including database writes (D1), object storage manipulation (R2), and outbound network requests via the
fetchAPI (SKILL.md). - Sanitization: The skill identifies the risk and recommends sanitizing user input as a core security best practice in the 'Security' section.
- [SAFE]: The skill uses well-known and trusted domains for its resources, including
developers.cloudflare.comandgithub.com/cloudflare. API configuration instructions rely on standard environment variable and local.envfile practices.
Audit Metadata