# Introduction

> PentAGI is a self-hosted, fully autonomous multi-agent system for penetration testing.

PentAGI is an open-source platform for automated security testing. You describe a target and a goal in plain language; a team of specialized AI [agents](https://pentagi.com/docs/glossary/#agent) plans the work, runs professional [tools](https://pentagi.com/docs/glossary/#tool) inside an isolated Docker [sandbox](https://pentagi.com/docs/glossary/#sandbox), remembers what it learned, and hands you a [report](https://pentagi.com/docs/glossary/#report).

It is built for security engineers, researchers and red teams who want an autonomous assistant they can run on their own infrastructure, with their own choice of language models.

* [Quick start](https://pentagi.com/docs/getting-started/quickstart/): Install PentAGI with the interactive installer.
* [Your first flow](https://pentagi.com/docs/getting-started/first-flow/): Run an autonomous flow and read its report.
* [Architecture](https://pentagi.com/docs/concepts/architecture/): Services, agents, memory and observability.
* [LLM providers](https://pentagi.com/docs/configuration/llm-providers/): Connect cloud, gateway or local models.

## [What it does](https://pentagi.com/docs/#what-it-does)

* **Autonomous testing.** The agents split each [task](https://pentagi.com/docs/glossary/#task) you give them into [subtasks](https://pentagi.com/docs/glossary/#subtask), choose the next step, execute it and adapt to the results — with optional execution monitoring and intelligent task planning for smaller models.
* **A team of specialists.** A primary agent delegates research, coding, installation and pentesting to specialized agents. Agents can ask an adviser for guidance, and with execution monitoring on, the adviser also reviews their work after a set number of [tool calls](https://pentagi.com/docs/glossary/#tool-call). A reflector steers any agent that stops calling tools back to a proper result.
* **Professional tooling.** 20+ tools such as nmap, metasploit and sqlmap run in a sandboxed container, built from a Docker image picked for each [flow](https://pentagi.com/docs/glossary/#flow).
* **Memory and knowledge.** Results and successful approaches are stored in PostgreSQL with pgvector for later reuse; an optional Neo4j [knowledge graph](https://pentagi.com/docs/glossary/#knowledge-graph) tracks relationships between hosts, services and [findings](https://pentagi.com/docs/glossary/#finding).
* **Web intelligence.** A built-in isolated browser and [search engines](https://pentagi.com/docs/glossary/#search-engine) — Tavily, Firecrawl, Traversaal, Perplexity, DuckDuckGo, Google, Sploitus and SearXNG — gather fresh information.
* **Observability.** Langfuse for LLM traces, and Grafana with OpenTelemetry, VictoriaMetrics, Jaeger and Loki for the whole system.
* **APIs.** REST and GraphQL with Bearer tokens for automation and integration.

## [Watch the overview](https://pentagi.com/docs/#watch-the-overview)

A short tour of PentAGI in action: [watch it on YouTube](https://youtu.be/R70x5Ddzs1o).

## [Current capability boundaries](https://pentagi.com/docs/#current-capability-boundaries)

* PentAGI is an autonomous and assistant-guided penetration testing platform, not a Breach and Attack Simulation product with predefined campaigns.
* Flow reports can be viewed on the web, copied, or downloaded as Markdown or PDF.
* [Providers](https://pentagi.com/docs/glossary/#provider) are available as built-in integrations and through custom OpenAI-compatible endpoints.

> [!WARNING]
> Only test systems you own or are explicitly authorized to assess. Use of PentAGI is subject to the [EULA](https://github.com/vxcontrol/pentagi/blob/main/EULA.md).