agency-agents: Detailed Introduction

[ikaijua]

agency-agents: Detailed Introduction

A collection of specialist roles, workflows, and deliverable requirements for existing AI tools, not a new foundation model.

• Audience: Users of Claude Code, Cursor, Codex, and similar tools who want reusable roles and task templates.
• Cost: The role library is MIT-licensed; host agents, models, and services are separate.

Table of Contents

Getting started

1. What it is and who it is for
2. Access and setup: your first session

Usage and configuration

3. A concrete task and how to check it
4. Costs: software, models, and execution
5. Configuration and extensions

Limits and references

6. Permissions, data, and limitations
7. Frequently asked questions
8. Official references

Use the contents to jump to each section. If an in-page link does not respond, open it in a new tab.

[ikaijua]

What it is and who it is for

In one sentence: A collection of specialist roles, workflows, and deliverable requirements for existing AI tools, not a new foundation model.

Who it suits: Users of Claude Code, Cursor, Codex, and similar tools who want reusable roles and task templates.

What you can do:

• Select roles for engineering, design, testing, or operations.
• Install workflows into tools you already use.
• Adapt instructions to your project.

Where to start: Start with the desktop app or repository. The app browses and installs roles.

Cost: The role library is MIT-licensed; host agents, models, and services are separate.

[ikaijua]

Access and setup: your first session

Start with the desktop app or repository. The app browses and installs roles.

For scripts, clone first and preview the scope:

git clone https://github.com/msitarzewski/agency-agents.git
cd agency-agents
./scripts/install.sh --tool opencode --division engineering --dry-run

Review destination and selection before installation; start with a small subset.

[ikaijua]

A concrete task and how to check it

Use one role for one task, such as read-only onboarding followed by a testing plan.

Example:

Use the installed frontend role to improve only the navigation component. Provide the diff, screenshots where tools permit, and checks actually run.

Role instructions do not automatically grant browser, screenshot, or test tools. Judge the delivered evidence.

[ikaijua]

Costs: software, models, and execution

Role Markdown and integration scripts are MIT-licensed. Execution still depends on the host’s subscription/API, compute, and connected tools.

Multiple roles can increase context and requests. First check whether a single role improves the task.

[ikaijua]

Configuration and extensions

The installer selects divisions or individual agents and supports multiple host formats. Discovery and role formats vary by tool.

Files include identity, workflow, deliverables, and evaluation criteria. Customize project constraints and acceptance criteria, not just names. Preserve customizations before updates.

[ikaijua]

Permissions, data, and limitations

Role templates do not create isolated employee accounts, permissions, or sandboxes and do not replace professional qualifications. The host controls execution access.

A large catalog is not a reason to install everything. Host registration limits and role-selection overhead make focused installs preferable.

[ikaijua]

Frequently asked questions

Is this a standalone agent runtime? Its core is a role and integration library; the app installs roles for a host.

Does a specialist name ensure expertise? No; the model, context, and tools still determine results.

What should I install? One or two directly relevant roles and useful deliverable templates.

[ikaijua]

Official references

• Role library
• Desktop app
• App releases
• License