higgsfield-product-photoshoot

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). The raw.githubusercontent.com link is dangerous because the skill instructs piping that install.sh into sh (curl ... | sh), which gives a remote script direct shell execution on the user host and is a classic malware distribution / remote code execution vector.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill contains high-risk supply-chain and data-exfiltration patterns: a remote "curl | sh" installer that enables arbitrary remote code execution and explicit automatic upload of local images to an external backend/CDN, which together could be abused to run malicious payloads or exfiltrate sensitive assets.

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow ingests outsider-authored free text directly from the user (the --prompt “short user-intent description” produced from interview answers) and submits it to Higgsfield’s backend for prompt enhancement and gpt_image_2 generation.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill includes a bootstrap instruction that downloads and executes a remote installer (curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh), which would run remote code to install the required Higgsfield CLI used at runtime.