yeet

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 parses repository content and user-generated templates from arbitrary GitHub repositories (e.g., "gh api repos/{owner}/{repo}/contents/.github/ISSUE_TEMPLATE" in references/create-issue.md and the GraphQL queries in references/create-discussion.md) and then uses that untrusted public content to choose templates, labels, categories, and to generate/modify issue/discussion/PR bodies—allowing third-party text to materially influence agent actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs gh api to fetch repository templates at runtime (e.g., gh api repos/{owner}/{repo}/contents/.github/ISSUE_TEMPLATE/{template_name} --jq '.content' | base64 -d), and the fetched YAML/Markdown templates are parsed and used to generate the issue/discussion body (i.e., they directly control the agent's prompts/instructions), so this is a high-confidence runtime external dependency.