> ## Documentation Index > Fetch the complete documentation index at: https://docs.neo.projectdiscovery.io/llms.txt > Use this file to discover all available pages before exploring further. # Overview > Specialized agents that power Neo tasks, and tools to build your own Neo runs every task through a set of specialized agents. When you start a task, Neo selects and coordinates the right agents automatically based on your objective. In Thorough mode, multiple agents run in parallel under the Agent Swarm orchestrator. You can also build custom agents with your own system context, tools, and configuration, and share them with your team or the community. ## Orchestration Agents that plan, coordinate, and manage the execution of other agents. Gathers intelligence and builds execution plans before action is taken Coordinates all agents in parallel for multi-phase assessments ## Discovery & Intelligence Agents that map attack surface, gather threat intelligence, and build context before active testing begins. Passive asset discovery, subdomain enumeration, and fingerprinting Asset inventory, discovery scans, and Nuclei templates from PDCP Security intelligence gathering and real-time threat research CVE research, exploit intelligence, and vulnerability database analysis Pre-execution reconnaissance and target intelligence for the Planner ## Testing & Exploitation Agents that actively test targets, execute exploits, and perform hands-on security assessments. Primary execution environment for active testing and the full security toolchain Real browser automation for DOM-based testing and authenticated flows REST and GraphQL API security testing with OpenAPI support Cross-site scripting detection, context analysis, and blind XSS testing Authorized adversary simulation and internal penetration testing Full Android pentest on cloud-hosted rooted devices with Frida instrumentation Advanced iOS testing with jailbroken device access and deep instrumentation Binary reverse engineering and vulnerability research On-demand container orchestration for spinning up test environments ## Analysis & Verification Agents that triage findings, verify exploitability, and ensure only confirmed issues reach your dashboard. Source-agnostic triage for HackerOne reports, scanner imports, and manual input Automated security-focused pull request review with inline findings Post-execution judge that independently re-tests findings and files confirmed issues Condenses large tool outputs into focused, security-relevant summaries