cli-anything-safari

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill exposes commands that accept and return secrets (e.g., form fields, cookies, set-cookie values) as command-line/JSON arguments and even shows plaintext password examples, so an agent using it would likely need to include secret values verbatim in its output.

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly exposes navigation and reading of arbitrary web pages (e.g., "cli-anything-safari --json tool navigate-and-read --url https://example.com" and "tool snapshot") and describes using that returned page content (refs, structured text) to drive clicks, fills, and script evaluation, so untrusted public web content can be ingested and materially influence agent actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The CLI explicitly downloads and runs the safari-mcp npm package at runtime (via npx, e.g. https://www.npmjs.com/package/safari-mcp / https://github.com/achiya-automation/safari-mcp), which fetches and executes remote code that the skill depends on.