distill

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

• Third-party content exposure detected (low risk: 0.10). The distill skill fetches and processes arbitrary web pages and files whose content is ingested by the agent and processed via LLM.

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

• Potentially malicious external URL detected (high risk: 0.70). The script invokes uvx with an explicit remote repository URL (https://github.com/grahama1970/treesitter-tools.git) to fetch, build, and execute external tools at runtime, constituting a weak/unknown operational dependency.