NYC

cloudflare-kv

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (HIGH): Automated security scans (URLite) identified a malicious phishing URL (keys.ma) associated with this skill. Phishing domains are used to harvest credentials or sensitive user information through deceptive redirects.\n- EXTERNAL_DOWNLOADS (MEDIUM): The skill depends on cloudflare-worker-base, which is an unverifiable third-party package. It is sourced from an untrusted repository (jezweb/claude-skills) that does not meet the Trusted External Sources criteria.\n- PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8).\n
  • Ingestion points: Processes untrusted data from KV storage and HTTP request bodies (e.g., getWithMetadata, req.text()) as described in README.md.\n
  • Boundary markers: No delimiters or isolation instructions are present to separate untrusted KV data from the agent's logic.\n
  • Capability inventory: The skill possesses extensive capabilities including reading, writing, and deleting data across the global Cloudflare edge network.\n
  • Sanitization: The provided operational examples lack any input validation, escaping, or filtering of data retrieved from external storage.
Recommendations
  • AI detected serious security threats
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 05:54 PM