vs-fix

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md explicitly instructs fetching GitHub issues with gh issue view <number> and extracting the issue body/steps to reproduce and expected/actual behavior (Phase 0 Step 1 and Phase 1), meaning it ingests untrusted, user-generated content from public GitHub issues which the agent will read and act on—allowing indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly calls the GitHub CLI to fetch issue content at runtime (e.g., via "gh issue view " which retrieves https://github.com///issues/), and that issue body/repro steps are extracted and injected into the agent's workflow, so remote GitHub issue content can directly control prompts.