scrapling

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

• Third-party content exposure detected (high risk: 0.85). At runtime, the skill’s required workflow can fetch arbitrary public web pages/HTML (e.g., via scrapling extract get/fetch/stealthy-fetch or MCP tools like get/fetch/stealthy_fetch), and that fetched page body is ingested as readable text/DOM into the agent’s LLM context (indirect prompt injection risk from outsider-authored page content).

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

• Potentially malicious external URL detected (high risk: 0.90). The script scripts/install.sh invokes pip install "scrapling[...]>=" which fetches remote package code from PyPI and then runs python -m scrapling.cli install (executing that installed remote code at runtime) — flagged command: pip install "scrapling[all]>=0.4.2".