CustomerFinderBot — agentic threat model

Not certain from the listing — likely utilizes commercial LLMs for NLP analysis and reply generation. The primary threat is indirect prompt injection, where adversarial social media posts manipulate the model's classification logic or hijack the reply generation process.

Not certain from the listing — ingests real-time data streams from X and Reddit. This creates a high exposure to data poisoning and malicious inputs embedded in public posts, which could lead to data exfiltration if the agent processes system prompts in the same context.

Not certain from the listing — likely uses a standard orchestration framework to manage scraping and LLM calls. Threats include insecure storage of platform API keys and lack of input sanitization before passing scraped content to the LLM.

Not certain from the listing — hosted as a closed-source SaaS. Key risks involve the secure storage of user credentials, API tokens for social media platforms, and potential container vulnerabilities in the hosting environment.

Not certain from the listing — no observability or guardrail mechanisms are mentioned. Without robust logging, prompt injection attempts or data drift in the social media scraping pipeline may go undetected.

Not certain from the listing — no compliance certifications (e.g., SOC2, GDPR) are specified. Compliance risks exist regarding the scraping and storage of personal data from social media users without explicit consent.

Not certain from the listing — operates as a standalone agent. The primary ecosystem risk is dependency on third-party APIs (X and Reddit), where changes in API policies or access controls could disrupt or compromise the agent's functionality.