heretic

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

• Third-party content exposure detected (medium risk: 0.65). Outsider free text can enter the LLM context via the discover mode’s live-web extraction routed through the scrapling skill, which fetches public web content at runtime and turns it into readable text for the agent.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's installer and usage explicitly fetch and install/clone remote code at runtime (e.g., pip install -U heretic-llm and git clone https://github.com/p-e-w/heretic and npx skills add https://github.com/akillness/jeo-skills), which downloads and then executes third‑party code required for the tool to run.