wrangler

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). SKILL.md instructs runtime retrieval from public Cloudflare documentation URLs (outsider-authored web content) and from a local schema file, but the required workflow explicitly includes fetching “Wrangler docs”/“Cloudflare docs” at runtime, which can inject free text into the agent’s LLM context.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs the agent to "Fetch the latest information" and lists external documentation URLs (https://developers.cloudflare.com/workers/wrangler/, https://developers.cloudflare.com/workers/, https://developers.cloudflare.com/workers/configuration/compatibility-dates/) that the agent is expected to retrieve at runtime to control CLI/config guidance, so these are runtime external dependencies that can directly influence prompts.