- Home
- Skills
- Agents & Orchestration
- Agent Fleet: AI Dev Team
Works with the AI Agents you already use
Agent Fleet: AI Dev Team
A complete multi-agent software development lifecycle for building scalable, documented, and architecturally sound products.
Free
Agent Fleet: AI Dev Team
Example session with this skill installed
Run the foundation phase for a new project called 'TaskTracker' based on the attached client brief and transcript.
- Read your context and instructions
- Compiled the agent fleet
I have initiated Phase A (Foundation) for the TaskTracker project. I have processed the client brief and transcript to establish the core requirements and initial module mapping.
Foundation Phase A: Initial Artifacts
| Phase | Agent | Deliverable | Status |
|---|---|---|---|
| F.0 | product-owner | docs/00-global/requirements.md | COMPLETED |
| F.1 | product-owner | docs/00-global/module-map.md | COMPLETED |
| F.2 | system-architect | docs/00-global/architecture.md | PENDING |
| F.3 | ux-engineer | docs/00-global/design-system.md | PENDING |
Module Build Queue
| Order | Module | Description | Dependency |
|---|---|---|---|
| 1 | auth-system | User registration, login, and JWT management | None |
| 2 | task-engine | Core CRUD for tasks and categories | auth-system |
| 3 | notification-service | Email and push alerts for task deadlines | task-engine |
| 4 | analytics-dashboard | User productivity reporting | task-engine |
Next steps
- Review the generated
docs/00-global/requirements.mdfor accuracy. - Approve the
module-map.mdbuild order. - Run
agent-fleet-foundationagain to proceed with Service Design (F.2).
Connects securely to your tools. The creator never sees your data.
About this skill
Building software in a multi-agent environment often leads to duplicated components, divergent API contracts, and fragmented documentation. The agent-fleet package provides a structured, multi-phase lifecycle for transforming client briefs into shipped software using a team of specialized agent roles.
What it does
- Foundation phase execution establishes project skeletons, global architecture, design systems, and shared data schemas before coding begins.
- Module loop orchestration manages the lifecycle of individual features from business requirements (BRD) to QA and closure.
- Parallel development synchronization uses frozen API contracts to allow frontend and backend agents to build, mock, and test in isolation.
- Inventory management tracks reusable components and services, preventing architectural drift and redundant implementations.
- Automated QA lifecycle generates test cases from requirements and executes black-box verification against running applications.
How it works
- Phase A (Foundation): Run
requirements-intakeandmodule-decompositionto define the project scope and build order. - Phase B (Module Loop): Execute
write-brdfollowed bygroom-modulefor the next item in the module map. - Execution: Use
api-contractto freeze the interface, then run frontend, backend, and QA tasks in parallel. - Integration: Run
swap-mocksande2e-verifyto ensure the module meets acceptance criteria. - Closure: Run
close-moduleto update the global inventory and changelog.
Frameworks & tools
This package is stack-agnostic and relies on a CONVENTIONS.md file to define specific technologies. It manages artifacts via Markdown templates and integrates with tools capable of launching subagents using specialized role definitions.
Why this beats prompting it yourself
Standard prompting often results in "one-shot" code that ignores global project state. This skill enforces a rigid phase-gate process where every agent must read the global inventory and adhere to frozen contracts, ensuring serial deliveries accumulate into a cohesive product rather than a collection of scripts.
Use cases
- Orchestrating a team of specialized agents to build a full-stack application.
- Maintaining architectural consistency across a large, multi-module project.
- Generating comprehensive technical documentation and test suites automatically.
- Scaling development by allowing parallel frontend and backend work without interface mismatch.
Known limitations
Requires a structured directory tree and specific templates to function. Initial setup of the foundation phase is mandatory before module-level work can begin.
Open source, also available as a Claude Code plugin: github.com/ahsuniqbal/agent-fleet
Requires pre-existing directory structure and templates. Phase A must be completed before starting Phase B modules. Logic depends on subagent orchestration capabilities.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 4 days ago
- Free with an Agensi account