company-creator

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

• Third-party content exposure detected (medium risk: 0.30). Skill.md’s required workflow for “Company From a Repo” instructs the agent to clone/read the user-provided git repo and scan/read README/package.json and discovered SKILL/AGENTS files at runtime, so outsider-authored free text can be ingested via the repo contents.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs running git against a remote repository at runtime (e.g., "git ls-remote https://github.com/owner/repo HEAD"), which means external repo content fetched from https://github.com/owner/repo can be read and used to construct agent instructions, creating a high-confidence runtime dependency.