cloudflare-worker

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection (MEDIUM): The skill provides patterns for processing untrusted external data which, when combined with the skill's file/database write capabilities, creates an attack surface for indirect prompt injection.
  • Ingestion points: The skill templates ingest untrusted data via c.req.json() and c.req.param() (hono.md), batch.messages (queues-testing.md), and request.body (storage.md).
  • Boundary markers: No explicit boundary markers or 'ignore embedded instructions' warnings are included in the provided code templates.
  • Capability inventory: The skill utilizes tools and code patterns with side-effect capabilities including env.DB.run (D1 SQLite write), env.STORAGE.put (R2 Object storage write), and the Bash tool for executing wrangler CLI commands.
  • Sanitization: While SQL patterns correctly use parameter binding, the R2 storage and queue processing patterns lack explicit sanitization or validation of untrusted input.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 11:36 AM