uv

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes an explicit example of passing an API token on the command line (uv publish --token pypi-xxx), which instructs embedding a secret verbatim in a command and therefore poses a high exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.80). The skill's documentation and workflow explicitly require fetching and resolving dependencies from external indexes and sources (e.g., SKILL.md "uv sync"/"uv add" and references/resolution.md "Sources" which list git URLs and arbitrary URL/index entries like httpx = { git = "https://github.com/encode/httpx" } and httpx = { url = "https://example.com/..." }), meaning the agent will ingest untrusted, user-controlled third‑party content that can materially affect resolution and runtime behavior.