LifeHubber.
Resources
AI Visuals
Agency Agents
Agency Agents is a large catalog of specialist instruction files for AI coding tools, covering roles across engineering, design, research, marketing, security, project management, and other fields.
You can browse and adapt individual roles, install selected agents or divisions with the project scripts, or use the companion desktop app to deploy and track them across supported tools and projects.
What it is
Reusable specialist instructions
Each role is a readable instruction file that sets a mission, working style, process, expected deliverables, and success measures. The files shape how an existing coding assistant approaches a task; they are not separate AI models or autonomous workers by themselves.
Why it stands out
One catalog, several coding tools
The repository includes converters and installers for Claude Code, Codex, Cursor, Gemini CLI, OpenCode, GitHub Copilot, Windsurf, Aider, and other documented tools. The companion app adds deterministic installs, project-level destinations, and a local record of what it wrote.
Availability
Public repository and desktop installer
The instruction catalog and scripts are available in a public MIT-licensed repository. A separate MIT-licensed companion app offers a graphical install, update, and reconciliation path for macOS, Linux, and Windows.
Why it matters
What makes it useful
A broad assistant often needs the same role, checklist, and output expectations explained again for each project. Agency Agents makes those working instructions reusable, so a builder can start from a specialist role, inspect every instruction, and adjust it to match the project instead of hiding the behavior inside a service.
What to know
Where it fits
Use the catalog when you want a coding assistant to approach a task as a focused reviewer, engineer, designer, researcher, or project specialist. Its Design Division includes an Image Prompt Engineer for subject, lighting and composition prompts, a UI Designer for components and prototypes, and a Visual Storyteller for storyboards and visual specifications. These are creative working instructions, not a built-in image generator. The value comes from the instructions and workflow expectations; results still depend on the underlying model, the surrounding tool, the context you provide, and the access you allow.
Notable points
What stands out
The repository supports both selective installation and full-catalog setup. The companion app records app-managed installs in a local ledger, compares rendered files to flag current, outdated, modified, removed, or foreign states, and backs up changed files before replacement or removal.
Before using
What to review
Read the chosen role file and remove instructions that do not fit your project, responsibilities, or risk level.
Install only the roles or divisions you need so the assistant has a clear job instead of a crowded roster.
Review installer and converter scripts before running them, especially when using a shell command that downloads or copies files.
The project's current release notes say the Windows desktop installer is not code-signed, so Windows SmartScreen may warn on first launch; use the official release page and review the notes before installing.
Check which files, tools, commands, accounts, and network destinations the host coding assistant can reach, and keep important actions behind human approval.
Check the original listing's documentation for current requirements, limitations, licensing and any relevant privacy or security considerations.
Reader fit
Who may find it relevant
Builders who want ready-made specialist roles they can read, edit, and reuse across AI coding tools.
Teams comparing how the same working role translates between assistants with different agent-file formats.
Less relevant if you want a finished autonomous agent platform that plans, runs, and monitors multi-agent work on its own.
Editorial note
Why LifeHubber lists it
Joining an unfamiliar codebase? The Codebase Onboarding Engineer role asks the assistant to trace a request or command through the actual files, explain the inputs and outputs, and identify what it has not inspected. Its instructions keep that task read-only, without refactoring suggestions. That gives you a starting role for understanding where a behavior lives before asking an assistant to change it; the instructions do not enforce file permissions.
Source links
Source materials
Reader note
About our listings
LifeHubber’s listings help you explore what’s available in AI and are not endorsements. We don’t independently verify every detail or guarantee that each listing is accurate, safe, actively maintained, or right for your needs.
What to explore next
Choose the role, then keep control of the work.
A specialist instruction file can sharpen one kind of task. Continue by comparing wider agent projects or checking what an agent may access before you install and use it.
More in AI Agents
Keep browsing this category
Explore more AI agent projects.
Paperclip
paperclipai/paperclip
A self-hosted server and dashboard for coordinating agent teams through companies, goals, roles, issues, heartbeats, budgets, approvals, and persistent activity records.
Storybook
storybookjs/storybook
Open-source UI workshop for previewing, documenting and testing component states in isolation, including changes made with coding assistants.
LocalAI
mudler/LocalAI
A self-hosted model service with compatible APIs, a web interface and selectable backends for language, speech and image generation, with model-specific hardware requirements.
For project maintainers
Listed here? You can use the badge.
If you maintain a project with a current LifeHubber listing, you may add the optional “Listed on LifeHubber AI Resources” badge to its README, docs, or website. No introduction or permission request is needed.
MORE ON LIFEHUBBER
Use AI on your own terms.
Explore more AI projects, stories, guides, tools and community insights across LifeHubber.
