# DapplePot

> DapplePot is an AI agent security and monitoring platform. It provides runtime security, session replay, and real-time threat detection for AI agent fleets — covering the OWASP LLM Top 10 and ASI Top 10.

DapplePot monitors, controls, and protects AI agents across their entire lifecycle: runtime interception, post-session replay, and cross-session pattern analysis. It surfaces prompt injection attacks, data leakage, excessive agency, and other AI-specific threats as they happen.

## Platform capabilities

- **Monitor**: Full session replay and real-time visibility into every AI agent action, tool call, and LLM exchange
- **Control**: Inline security checks that can sanitize content, block calls, or terminate sessions on policy violations
- **Protect**: Covers OWASP LLM Top 10 and ASI Top 10 threat categories out of the box

## Who it is for

- **Security teams**: Detect and respond to AI agent threats across the fleet
- **AI/ML teams**: Observe agent behaviour, debug sessions, and validate safety in production
- **Audit & compliance teams**: Session logs and policy enforcement for regulatory requirements

## Developer SDK

The DapplePot Python SDK instruments AI agents built with Anthropic, OpenAI, LangChain, or LangGraph. Full documentation at https://docs.dapplepot.com.

Install: `pip install dapplepot-sdk`

## Links

- Website: https://www.dapplepot.com
- Developer docs: https://docs.dapplepot.com