pp-botsee

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow runs the botsee-pp-cli commands (e.g., ai-visibility-audit, site-config, sites-summary) which fetch and return site/analysis data and LLM response text from BotSee/API into the agent via the CLI’s --agent JSON stdout, so any outsider-authored web content or synced third-party LLM responses that BotSee includes can become LLM-readable context (indirect prompt injection risk).

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

• Potentially malicious external URL detected (high risk: 0.90). The fallback install step runs "go install github.com/mvanhorn/printing-press-library/library/marketing/botsee/cmd/botsee-pp-cli@latest", which fetches remote source and compiles/executes code at runtime, making it a high-risk external dependency.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill exposes payment-related commands: it can create credit-card signup tokens ("signup via-cc") and USDC signup tokens ("signup via-usdc-token"), and it includes billing management such as "billing update-settings" (update the organization's monthly spend limit) and access to auto-charge/transaction history. These are explicit financial/payment APIs (card + USDC signup tokens and billing settings), not generic tooling — therefore it grants direct financial execution/management capabilities.