frontend-landings

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

• Third-party content exposure detected (high risk: 0.90). The skill's "Improve / Convert" workflow in SKILL.md explicitly accepts a URL as input ("User provides existing HTML, a description, a screenshot, or a URL → enhance") and requires the agent to analyze the page's structure and regenerate HTML, which means it would fetch and interpret untrusted public web content that can materially influence subsequent generation and actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The README contains installer commands that fetch and run remote code and skill files—e.g. the one-line installer curl -sSL https://raw.githubusercontent.com/auwalmusa/frontend-landings-claude-skill/main/install.sh | bash (executes remote script) and curl downloads of https://raw.githubusercontent.com/auwalmusa/frontend-landings-claude-skill/main/SKILL.md and https://raw.githubusercontent.com/auwalmusa/frontend-landings-claude-skill/main/references/STYLE_PRESETS.md (which, when installed, directly control the agent's prompts/instructions)—so these URLs are runtime-fetched dependencies that can execute code or control agent behavior.