Skip to content

Examples

Agent Teams Module ships with three bundled example briefs that demonstrate different project types. All are in examples/.

Example 1: Software Project (examples/software-project/)

A FastAPI backend with authentication and a task management API.

Brief summary:

{
  "project_name": "WebAppBackend",
  "project_goal": "Build a Python FastAPI backend for a task management web application, including REST API endpoints, database models, authentication, and automated tests.",
  "deliverables": ["Python modules", "OpenAPI documentation", "test suite"],
  "output_format": "Python 3.11 modules",
  "primary_output_dir": "src/"
}

What gets generated:

  • Orchestrator + 10 governance agents
  • Domain agents: @primary-producer, @quality-auditor, @technical-validator, @format-converter
  • Workstream experts: one per component (e.g. @auth-module-expert, @tasks-api-expert)
  • Specialist agents for FastAPI and PostgreSQL
  • copilot-instructions.md wired to the full team

To run this example:

agentteams \
  --description examples/software-project/brief.json \
  --project /tmp/webappbackend \
  --framework copilot-vscode

The expected/ directory shows the agent files this brief produces.

Example 2: Data Pipeline (examples/data-pipeline/)

An ETL pipeline that ingests daily sales CSVs, transforms and loads them into a PostgreSQL warehouse, and generates weekly PDF reports.

Brief summary:

{
  "project_name": "SalesDataPipeline",
  "project_goal": "Build an ETL pipeline that ingests daily sales CSV exports, validates and transforms them, loads them into a PostgreSQL warehouse, and produces weekly summary reports.",
  "deliverables": ["Python ETL modules", "SQL transformation scripts", "weekly PDF reports"],
  "output_format": "Python 3.11 modules and PDF reports",
  "primary_output_dir": "src/"
}

What gets generated:

  • Full governance + domain agent set
  • Specialist agents for PostgreSQL and the ETL build system
  • Workstream experts scoped to each pipeline stage
  • Reference files for pandas and SQLAlchemy

To run this example:

agentteams \
  --description examples/data-pipeline/brief.json \
  --project /tmp/salespipeline \
  --framework copilot-vscode

Example 3: Research Project (examples/research-project/)

A peer-reviewed academic paper on multi-agent coordination theory, progressing from an HTML outline through a final LaTeX manuscript.

Brief summary:

{
  "project_name": "ResearchPaperProject",
  "project_goal": "Produce a peer-reviewed academic paper on multi-agent coordination theory, progressing from an outline through final LaTeX manuscript.",
  "deliverables": ["HTML chapter drafts", "LaTeX manuscript", "BibTeX bibliography"],
  "output_format": "PDF via LaTeX",
  "primary_output_dir": "html/chapters/"
}

What gets generated:

  • Domain agents include @reference-manager, @format-converter, and @quality-auditor
  • Authority hierarchy sourced from academic papers and agent documentation
  • Workstream experts per chapter / manuscript section
  • copilot-instructions.md tuned for academic writing conventions

To run this example:

agentteams \
  --description examples/research-project/brief.json \
  --project /tmp/researchpaper \
  --framework copilot-vscode

Writing Your Own Brief

See Description Format for the full brief.json schema reference, or browse the JSON schema directly.

The minimum viable brief requires only three fields:

{
  "project_name": "MyProject",
  "project_goal": "one sentence describing what you are building",
  "deliverables": ["list of output types"]
}

All other fields are auto-inferred or prompted via SETUP-REQUIRED.md after generation.