coregit

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

• Third-party content exposure detected (high risk: 1.00). The skill directly reads and indexes arbitrary user-generated repository content and LLM Wiki sources (e.g., GET /v1/repos/:slug/blob/:ref/:path, the wiki sources/pages endpoints, /wiki/llms.txt and the repo's schema.md) which are untrusted third-party files that can contain agent instructions and thus materially influence agent decisions and tool use.

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

• Potentially malicious external URL detected (high risk: 0.90). The onboarding step runs npx coregit-wizard@latest (which fetches and executes remote package code from the npm registry, e.g. https://registry.npmjs.org/coregit-wizard), so this external code is fetched and executed at runtime and is relied on for creating the API key/account.