pp-daraz

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The document exposes explicit data-exfiltration features (a webhook delivery sink and an optional automatic remote feedback POST) that can be used by an agent to send command output or local feedback to external endpoints without interactive confirmation.

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md describes a CLI that “drives…scraper” behavior (searches Daraz.pk listings and reads reviews at runtime), so the agent will likely ingest outsider-authored review/listing text from the public web into LLM-visible JSON/stdout.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill's prerequisite install steps call go install on remote repositories (github.com/mvanhorn/printing-press-library/library/commerce/daraz/cmd/daraz-pp-cli@latest and github.com/mvanhorn/printing-press-library/library/commerce/daraz/cmd/daraz-pp-mcp@latest), which fetch and install remote code during setup and are required for the skill to run.