cloudflare-r2

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (LOW): The skill utilizes npx wrangler for bucket and object operations, which involves subprocess execution of an external tool.
  • [EXTERNAL_DOWNLOADS] (LOW): Using npx results in the download and execution of the wrangler package from the npm registry. Since the package is unversioned and Cloudflare is not on the specific 'Trusted External Sources' list, this is a minor dependency risk.
  • [PROMPT_INJECTION] (LOW): An indirect prompt injection surface exists through the storage of user-controlled files. 1. Ingestion points: User-provided file paths and contents handled by the object put command. 2. Boundary markers: No delimiters or ignore instructions are present. 3. Capability inventory: CLI execution via npx. 4. Sanitization: No sanitization of file paths or contents is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 07:34 PM