What is Pentagi?

Pentagi is an AI-powered autonomous penetration testing platform that executes complex tests via terminal, browser and editor automation.It uses a RAG-enabled knowledge graph (Graphiti/Neo4j) and semantic memory to track context and inform attack paths, while command history and outputs are stored in PostgreSQL.

Pentagi packages 200+ Kali Linux tools in optimized Docker images and runs tests sandboxed for self-hosted deployment and infrastructure control.Users can monitor activities in real time through a web UI and observability stack (Grafana, Loki, Jaeger, OpenTelemetry, ClickHouse, VictoriaMetrics) and export structured results and reports.

The platform supports multiple LLM providers (OpenAI, Anthropic, Google Gemini, AWS Bedrock, Ollama and others) and exposes REST and GraphQL APIs for integration and automation.Target users include security researchers, penetration testers, red teams and DevOps teams seeking automated test execution, reproducible workflows, and extensible integrations.

Pentagi user reviews

Would you recommend Pentagi?

Recommend this tool?

Pentagi's key features

Autonomous penetration testing via terminal, browser, and editor automation
RAG-enabled knowledge graph (Graphiti/Neo4j) with semantic memory for context tracking and attack path reasoning
Packaged Kali Linux tools in optimized Docker images for sandboxed self-hosted test execution
Real-time monitoring through a web UI and observability stack (Grafana, Loki, Jaeger, OpenTelemetry, ClickHouse, VictoriaMetrics) with exportable structured results and reports
Support for multiple LLM providers (OpenAI, Anthropic, Google Gemini, AWS Bedrock, Ollama, etc.) and integration via REST and GraphQL APIs

Pentagi use cases

Automate continuous security assessments of web applications and APIs using Pentagi's LLM-powered autonomous pentest automation and reproducible automated workflows, executing terminal, browser and editor attacks in a sandboxed Docker environment with 200+ Kali tools and monitoring results on the real-time dashboard
Integrate Pentagi into CI/CD pipelines via API-driven pentest integration to run pre-production vulnerability scans, leveraging the RAG-enabled knowledge graph and semantic memory to plan targeted attack paths and prioritize remediation across releases
Conduct internal red-team exercises and compliance audits in a self-hosted deployment, using Pentagi's containerized pentest sandbox to safely reproduce exploits, bundle findings into exportable reports for auditors and reuse attack steps with semantic memory for consistent retesting