# The 12 Named Agents in PAI: Specializations and Roles > Explore the 12 named agents in Personal AI Infrastructure PAI. Discover their unique specializations from Algorithm and Architect to research specialists and understand their crucial roles in system design and security. - Repository: [Daniel Miessler 🛡️/Personal_AI_Infrastructure](https://github.com/danielmiessler/personal_ai_infrastructure) - Tags: deep-dive - Published: 2026-02-19 --- **The Personal AI Infrastructure (PAI) framework defines 12+ specialized named agents—including Algorithm, Architect, Engineer, Artist, Designer, QATester, Pentester, Intern, and five research specialists—that each handle distinct tasks from system design to security testing and multi-perspective research.** The `danielmiessler/Personal_AI_Infrastructure` repository provides a structured approach to AI-assisted development through specialized agent personalities. These **12 named agents in PAI** are defined in the Named Agents section of the v3.0 README, located at [`Releases/v3.0/README.md`](https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Releases/v3.0/README.md) (lines 65-79), and provide focused expertise for complex software engineering workflows. ## Core Development Agents The PAI framework includes eight core development agents that cover the software development lifecycle from design to deployment. ### Algorithm Agent The **Algorithm** agent specializes in ISC tracking, verification, constraint extraction, and PRD management. This agent handles complex logic validation and ensures algorithmic constraints are properly documented and tested. ### Architect Agent The **Architect** agent focuses on system design, distributed systems, and infrastructure planning. It provides high-level structural guidance for scalable application architecture. ### Engineer Agent The **Engineer** agent handles test-driven development, implementation patterns, and brings Fortune-10 level engineering experience to code generation. This agent specializes in production-ready code implementation. ### Artist Agent The **Artist** agent manages visual content creation, prompt engineering, and model selection for generative art. It optimizes visual outputs through careful prompt crafting and model choice. ### Designer Agent The **Designer** agent is responsible for UX/UI design, accessibility compliance, and shadcn/ui component implementation. This agent ensures interfaces meet modern design standards and accessibility requirements. ### QATester Agent The **QATester** agent specializes in browser automation, verification workflows, and test-coverage generation. It automates quality assurance processes across web applications. ### Pentester Agent The **Pentester** agent focuses on security testing and vulnerability assessment. This agent identifies security weaknesses and validates defensive measures. ### Intern Agent The **Intern** agent acts as a high-agency generalist capable of multi-tool problem solving across domains. This agent handles miscellaneous tasks that require flexibility and broad tool integration. ## Research Specialist Agents Beyond core development, PAI includes five specialized research agents that provide multi-perspective analysis through different cognitive frameworks. ### GeminiResearcher The **GeminiResearcher** conducts multi-perspective parallel investigations, exploring topics from multiple angles simultaneously to provide comprehensive coverage. ### GrokResearcher The **GrokResearcher** provides contrarian, fact-based analysis. This agent challenges assumptions and verifies claims through skeptical inquiry. ### CodexResearcher The **CodexResearcher** performs technical archaeology and curiosity-driven probing of codebases. This agent specializes in understanding legacy systems and deep technical exploration. ### ClaudeResearcher The **ClaudeResearcher** specializes in academic synthesis and scholarly source aggregation. This agent excels at compiling and summarizing academic literature and formal research. ### PerplexityResearcher The **PerplexityResearcher** handles real-time web search and citation-heavy research. This agent provides current information with extensive source verification. ## How to Reference PAI Agents in Practice When working with the Personal AI Infrastructure, you invoke specific agents by name to leverage their specialized capabilities. The agent definitions are located in the repository's versioned release documentation. ```text # Example agent invocation pattern @Algorithm: Verify constraint extraction for this PRD @Architect: Design distributed system for microservices @Pentester: Assess authentication vulnerabilities @ClaudeResearcher: Synthesize academic sources on distributed systems ``` Each agent maintains distinct system prompts and operational parameters defined in [`Releases/v3.0/README.md`](https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Releases/v3.0/README.md), ensuring consistent behavior across sessions. ## Summary The Personal AI Infrastructure provides **12+ named agents in PAI** that cover comprehensive software development and research needs: - **Core Development**: Algorithm, Architect, Engineer, Artist, Designer, QATester, Pentester, and Intern handle implementation, design, testing, and security. - **Research Specialists**: GeminiResearcher, GrokResearcher, CodexResearcher, ClaudeResearcher, and PerplexityResearcher provide multi-perspective analysis and technical investigation. - **Source Location**: All agent definitions are documented in [`Releases/v3.0/README.md`](https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Releases/v3.0/README.md) (lines 65-79) within the `danielmiessler/Personal_AI_Infrastructure` repository. ## Frequently Asked Questions ### What are the 12 named agents in PAI? The 12 core named agents in PAI include Algorithm, Architect, Engineer, Artist, Designer, QATester, Pentester, and Intern for development tasks, plus four research specialists. However, the framework actually defines 13 total agents including five research specialists (GeminiResearcher, GrokResearcher, CodexResearcher, ClaudeResearcher, and PerplexityResearcher), as documented in the v3.0 README. ### How do the research agents in PAI differ from each other? Each research agent in PAI employs a distinct cognitive approach. **GeminiResearcher** conducts parallel multi-perspective investigations, **GrokResearcher** provides contrarian fact-checking, **CodexResearcher** performs deep technical archaeology, **ClaudeResearcher** synthesizes academic sources, and **PerplexityResearcher** executes real-time web searches with heavy citation. ### Where are the PAI agent specializations defined in the source code? The agent specializations are defined in the **Named Agents** section of the PAI v3.0 README, specifically located at [`Releases/v3.0/README.md`](https://github.com/danielmiessler/Personal_AI_Infrastructure/blob/main/Releases/v3.0/README.md) between lines 65-79 in the `danielmiessler/Personal_AI_Infrastructure` repository. This documentation specifies each agent's role, expertise areas, and operational focus. ### What is the difference between the Engineer and Architect agents in PAI? The **Architect** agent specializes in high-level system design, distributed systems planning, and infrastructure architecture, focusing on structural decisions and scalability. The **Engineer** agent handles implementation details, test-driven development, coding patterns, and production-ready code generation, bringing Fortune-10 level engineering practices to execution tasks.