wrangler-coder
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for Cloudflare Workers that process untrusted data from external HTTP requests, which is the standard operational model for such agents.
- Ingestion points: Worker scripts in
references/d1-r2.mdandreferences/worker-example.mdingest untrusted data throughrequest.body,request.json(), andurl.pathname. - Boundary markers: The Workers AI examples in
references/workers-ai.mddo not include specific boundary markers or instructions to the model to ignore potential injection within the prompt data. - Capability inventory: The worker templates include write access to D1 (SQL), R2 (Storage), and KV (Key-Value) bindings, and the ability to execute Workers AI tasks.
- Sanitization: The skill mitigates common injection risks by using parameter binding (
.bind()) for SQL queries in both the D1 and general worker examples, which is a key security best practice.
Audit Metadata