Skip to content

QUICKSTART.md

Latest commit

 

History

History
114 lines (82 loc) · 2.97 KB

QUICKSTART.md

File metadata and controls

114 lines (82 loc) · 2.97 KB

Quickstart Guide

Get started with the Agent Orchestrator MVP in 3 minutes.

Prerequisites

Setup

1. Install dependencies

pip install -r requirements.txt

2. Configure API key

Create a .env file (or export environment variable):

cp .env.example .env
# Edit .env and add your API key

Or export directly:

export ANTHROPIC_API_KEY="your-api-key-here"

3. Run your first agent

python test_agent.py requirements_gatherer

This will:

  • Load the Requirements Gatherer agent prompt from prompts/requirements_gatherer.md
  • Execute it via Claude API with a sample finance CLI project description
  • Save output to workflow/artifacts/requirements_gatherer/
  • Display formatted results

What Just Happened?

The orchestrator:

  1. Loaded the agent prompt - Read the full agent identity and instructions from the prompt file
  2. Assembled context - Prepared project context (project name, current phase)
  3. Called Claude API - Sent the prompt + context + user task to Claude
  4. Saved output - Stored the agent's response as an artifact
  5. Recorded execution - Logged the execution in workflow/executions.json

Output Location

workflow/
├── artifacts/
│   └── requirements_gatherer/
│       └── requirements_gatherer_output.md  # Agent output
├── executions.json                           # Execution history
└── current_state.json                        # Workflow state

Try Other Agents

# Run system architect (requires requirements_gatherer output as context)
python test_agent.py system_architect

# Run any agent by ID
python test_agent.py <agent_id>

See config/agents.yaml for the full list of 31 available agents.

MVP Scope

This MVP demonstrates:

  • ✅ Loading agent prompts from markdown files
  • ✅ Executing single agents via Claude API
  • ✅ Simple context assembly
  • ✅ Artifact storage and retrieval
  • ✅ Execution history tracking

Not yet implemented (future):

  • ❌ Full DAG-based workflow orchestration
  • ❌ Parallel agent execution
  • ❌ Context Manager with SQLite
  • ❌ Output validation schemas
  • ❌ Approval gates
  • ❌ Error recovery with Debugger agent
  • ❌ Kanban task management

Next Steps

  1. Review agent output in workflow/artifacts/
  2. Read the agent prompts in prompts/ to understand each agent's role
  3. Check CLAUDE.md for the full orchestration architecture
  4. Extend the MVP to support multi-agent workflows

Troubleshooting

"ANTHROPIC_API_KEY not found"

  • Make sure you've set the environment variable or created a .env file

"Agent not found in configuration"

  • Check config/agents.yaml for valid agent IDs
  • Agent IDs use underscores (e.g., requirements_gatherer, not requirements-gatherer)

"Prompt file not found"

  • Verify the prompt file exists in prompts/
  • Check the prompt_file path in config/agents.yaml