Agency Agents is a complete AI agency collection, bringing together more than 230 carefully crafted specialized AI agent personas, distributed across engineering, design, paid media, sales, marketing, product, project management, testing, security, support, spatial computing, specialized fields, finance, game development, academia, geospatial, healthcare, research, and many other departments. Each agent has a unique identity and personality, a core mission and workflow, technical deliverables with code examples, success metrics, and a communication style, emphasizing specialization, deliverability, and production-environment validation.
The project can be used flexibly. The recommended approach is to install the native desktop application (supporting macOS, Linux, and Windows), which allows browsing the full agent roster and installing to tools such as Claude Code, Cursor, Codex, and Gemini with one click, and updates automatically. It can also be installed via command-line scripts, or used directly as reference material to browse, copy, and adapt the needed agent files. Supported tools include Claude Code, GitHub Copilot, Antigravity, Gemini CLI, OpenCode, Cursor, Aider, Windsurf, OpenClaw, Qwen Code, Kimi Code, Codex, Osaurus, Hermes, Mistral Vibe, DeepSeek Harness, and others, with accompanying conversion and installation scripts.
During installation, integration files can be generated via scripts, then an interactive installer can be run to automatically detect tools already installed on the system, select the agents to install via checkboxes, or install directly for a specific tool, or use non-interactive mode for CI/scripts. On multi-core machines, parallel processing can be enabled to speed up installation. Users can selectively install by department or by individual agent, for example installing only the teams they need.
Applicable scenarios are broad, for example building a startup MVP, launching a marketing campaign, enterprise-level feature development, paid media account takeover, full-agency product discovery, smart campus digital twins, and more. Each scenario demonstrates how to combine agents from different departments to work together, producing a complete solution from technical architecture to market validation. The project welcomes contributions, including adding new agents, improving existing agents, and sharing use cases, and follows a template structure containing front matter information, identity and memory, core mission, key rules, technical deliverables, workflow, and success metrics.
Unlike general AI prompts or prompt libraries, Agency Agents provides a deeply specialized agent system with personality and processes, transparent, branchable, adaptable, and already battle-tested in production environments. The project originated from a Reddit discussion, and after months of iteration, contains more than 10,000 lines of personas, processes, and code examples. Regarding fees, pricing, specific platform support details, etc., the official text does not provide further information.
