fastlike

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

• Third-party content exposure detected (high risk: 0.70). The skill's SKILL.md explicitly instructs the agent to clone and read the fastlike source from the public GitHub repo (https://github.com/avidal/fastlike and files under ~/src/fastlike) and requires reading those third-party source files as the authoritative reference for answers, which means the agent will ingest and act on untrusted, public repository content.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent to clone and read the fastlike repository at https://github.com/avidal/fastlike (git clone https://github.com/avidal/fastlike.git ~/src/fastlike) at runtime so that that remote source controls answers, making it a required external dependency that can directly influence prompts/responses.