maintaining-a-marketplace

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs adding marketplaces from public git locations (e.g., "Users add your marketplace by its git location" and "/plugin marketplace add owner/repo" and "source": { "source": "github", ... }) and requires reading plugin.json from those repositories ("Read the plugin's plugin.json to extract metadata"), meaning the agent fetches and interprets untrusted third-party repository content which can influence installation/versioning decisions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill's runtime commands include adding marketplaces from git (e.g., "/plugin marketplace add https://gitlab.com/company/plugins.git"), which will fetch remote repository content that can contain plugins/agents/skills that directly control prompts or execute code, so this is a runtime external dependency.