Agent Templates — Tech Team

Ready-to-use agent templates for building a complete AI tech team. Each folder contains the skills, identity, and personality for one specialized agent.

Full writeup: https://dev.to/zvone187/i-built-an-autonomous-dev-team-with-3-ai-agents-that-takes-a-linear-ticket-all-the-way-to-a-pull-291g

What’s Inside

Each folder contains:

• IDENTITY.md — The agent’s name, role, and personality
• SOUL.md — Core values, behavioral guidelines, and working style
• Skill folders — Each skill has a SKILL.md with full instructions the agent follows
• environment.md (QA only) — Configuration template for test infrastructure

How to Set Up

These templates work with OpenClaw or Pazi agents. You can set up each agent individually or all three as a team.

Option 1: Set Up Each Agent Individually

For each agent, create a new agent in your workspace, install the skills, and run the setup wizard.

---

🔨 Developer Agent

1. Upload or share the entire Developer Agent folder with your agent
2. Send this prompt:

Here is a template for a Developer Agent. Please:
1. Create a new agent called "Developer Agent"
2. Install all skills from this folder into the new agent's workspace
3. Set up the agent's IDENTITY.md and SOUL.md from the files provided
4. Load the "Build Developer Agent" skill and kick off the setup process

The setup wizard will walk us through connecting GitHub, Linear, coding agents (Codex/Claude Code), and verifying everything works.

---

🧪 QA Agent

1. Upload or share the entire QA Agent folder with your agent
2. Send this prompt:

Here is a template for a QA Agent. Please:
1. Create a new agent called "QA Agent"
2. Install all skills from this folder into the new agent's workspace
3. Set up the agent's IDENTITY.md, SOUL.md, and environment.md from the files provided
4. Load the "Build QA Agent" skill and kick off the setup process

The setup wizard will walk us through connecting GitHub, analyzing the repo architecture, discovering infrastructure, and configuring environment.md with all the values the QA skills need.

---

🎯 Tech Lead Agent

1. Upload or share the entire Tech Lead Agent folder with your agent
2. Send this prompt:

Here is a template for a Tech Lead Agent. Please:
1. Create a new agent called "Tech Lead Agent"
2. Install all skills from this folder into the new agent's workspace
3. Set up the agent's IDENTITY.md and SOUL.md from the files provided
4. Load the "Build Tech Lead Agent" skill and kick off the setup process

The setup wizard will walk us through connecting Linear, Slack, and configuring the agent to orchestrate Developer and QA agents for ticket management.

---

Option 2: Set Up the Full Tech Team

If you want all three agents working together, set them up in this order:

1. Developer Agent first — this is your primary coding agent
2. QA Agent second — handles testing for everything the Developer builds
3. Tech Lead Agent last — orchestrates the other two, needs their agent IDs during setup

The Tech Lead’s setup wizard will ask for the Developer and QA agent IDs so it can assign and coordinate work between them.

How It Works

Once set up, the team operates like this:

1. Tech Lead receives a Linear ticket and creates a plan
2. Tech Lead assigns implementation to the Developer
3. Developer runs cross-review planning (Codex + Claude Code), implements the feature, creates a PR
4. Tech Lead assigns testing to QA
5. QA runs the full test suite with browser/API tests and screenshot evidence
6. Tech Lead reports results back — PR is ready for human review and merge

Humans stay in the loop for PR merges and final decisions. The agents handle the execution.

Skill Overview

Requirements

• OpenClaw or Pazi agent runtime
• API keys for your LLM provider (Anthropic, OpenAI, etc.)
• GitHub access for the repos you’re working on
• Linear access (optional but recommended for ticket management)

License

MIT