iterate-from-user

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and ingests open/public third-party content — e.g., article URLs via WebFetch/WebSearch and GitHub issues via gh issue view (see the "Branch A — article-link" and "Branch B — resource-link" sections of SKILL.md) — and the agent must read and synthesize that untrusted/user-generated content into proposals that determine next actions.

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

• Potentially malicious external URL detected (high risk: 0.90). This skill explicitly fetches user-supplied external content at runtime (e.g., article URLs via WebFetch and GitHub issue URLs such as https://github.com///issues/), and injects that fetched content into the agent's context to directly drive prompts/instructions for producing the Proposal, so the external URL content is a required runtime dependency that can control agent behavior.