Skip to content

cli.md

Latest commit

 

History

History
331 lines (262 loc) · 6.65 KB

cli.md

File metadata and controls

331 lines (262 loc) · 6.65 KB

CLI Usage Guide

Overview

The Code-Swarm CLI provides commands for managing agents, tasks, and deployments.

Commands

Server Management

Start API Server

code-swarm api [OPTIONS]

Options:
  --host TEXT          Host to bind to (default: 127.0.0.1)
  --port INTEGER       Port to bind to (default: 8000)
  --reload             Enable auto-reload on code changes
  --workers INTEGER    Number of worker processes (default: 1)

Example:

code-swarm api --host 0.0.0.0 --port 8000 --reload

Agent Management

List Agents

code-swarm agents list [OPTIONS]

Options:
  --role TEXT          Filter by role (backend, frontend, etc.)
  --status TEXT        Filter by status (idle, active, completed)
  --format TEXT        Output format: json, table (default: table)

Example:

code-swarm agents list --role backend --format json

Output:

┌─────────┬─────────┬─────────┬──────────┐
│ ID      │ Name    │ Role    │ Status   │
├─────────┼─────────┼─────────┼──────────┤
│ agent_1 │ solver  │ backend │ idle     │
│ agent_2 │ designer│ frontend│ active   │
└─────────┴─────────┴─────────┴──────────┘

Create Agent

code-swarm agents create [OPTIONS]

Options:
  --name TEXT          Agent name (required)
  --model TEXT         Model ID (default: gpt-4)
  --role TEXT          Agent role (backend, frontend, etc.)
  --capabilities TEXT  Comma-separated capabilities

Example:

code-swarm agents create \
  --name solver \
  --role backend \
  --capabilities code-generation,testing

Show Agent Details

code-swarm agents show <agent_id>

Output:

Agent: solver (agent_1)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Model:         gpt-4
Role:          backend
Status:        idle
Capabilities:  code-generation, testing
Created:       2026-05-03 10:00:00 UTC

Task Management

List Tasks

code-swarm tasks list [OPTIONS]

Options:
  --status TEXT        Filter by status (pending, active, completed)
  --assigned-to TEXT   Filter by agent
  --format TEXT        Output format: json, table

Example:

code-swarm tasks list --status pending

Create Task

code-swarm tasks create [OPTIONS]

Options:
  --title TEXT         Task title (required)
  --description TEXT   Detailed description
  --priority INTEGER   Priority level 1-10
  --assigned-to TEXT   Agent to assign task to

Example:

code-swarm tasks create \
  --title "Fix login endpoint" \
  --description "Add JWT refresh token" \
  --priority 8 \
  --assigned-to solver

Show Task Details & Logs

code-swarm tasks show <task_id>

Output:

Task: Fix login endpoint (task_1)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Status:       active
Priority:     8/10
Assigned:     solver
Created:      2026-05-03 10:00:00 UTC
Progress:     ████████░░ 80%

Last 5 log entries:
  [10:05:20] Analyzing endpoint...
  [10:05:25] Found 3 issues
  [10:05:30] Generating fix...
  [10:06:15] Testing fix...
  [10:06:45] Fix validated ✓

Watch Task Execution (Real-time)

code-swarm tasks watch <task_id>

Flags:
  --follow     Continue watching until completion
  --filter     Filter logs by agent or action

Example:

code-swarm tasks watch task_1 --follow

Output (streaming):

task_1: hermes | Distributing task...
task_1: prometheus | Analyzing requirements...
task_1: zeus | Planning architecture...
task_1: atlas | Generating backend...
task_1: iris | Generating frontend...
task_1: [COMPLETE] ✓ All systems ready

Deployment

Deploy to Vercel

code-swarm deploy vercel [OPTIONS]

Options:
  --project TEXT       Vercel project name
  --env-file TEXT      Path to .env file
  --production         Deploy to production

Example:

code-swarm deploy vercel --project code-swarm --production

Deploy to Kubernetes

code-swarm deploy k8s [OPTIONS]

Options:
  --context TEXT       Kubernetes context
  --namespace TEXT     Target namespace
  --replicas INTEGER   Number of replicas

Database

Initialize Database

code-swarm db init

Creates tables and indices in Supabase.

Run Migrations

code-swarm db migrate

Applies pending schema migrations.

Monitoring

Show System Status

code-swarm status

Output:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🟢 API Server: healthy
🟢 Database: connected
🟢 Simone-MCP: connected
🟡 Cache: degraded

Active Agents: 3
Running Tasks: 12
Avg Response Time: 145ms

Show Logs

code-swarm logs [OPTIONS]

Options:
  --tail INTEGER       Show last N lines (default: 50)
  --agent TEXT         Filter by agent
  --level TEXT         Log level (debug, info, warning, error)
  --follow             Follow logs in real-time

Example:

code-swarm logs --agent solver --level error --follow

Configuration

Show Config

code-swarm config show

Output:
SUPABASE_URL: postgresql://...
SIMONE_MCP_URL: http://your-simone-host:8234
PRIMARY_MODEL: gpt-4
ENVIRONMENT: production

Edit Config

code-swarm config set <KEY> <VALUE>

Example:
code-swarm config set PRIMARY_MODEL gpt-4-turbo

Environment Variables

The CLI respects all environment variables defined in .env:

export SUPABASE_URL=postgresql://...
export SIMONE_MCP_URL=http://your-simone-host:8234
export PRIMARY_MODEL=gpt-4
export SECRET_KEY=your-secret-key

Tips & Tricks

Batch Operations

# Create 5 agents at once
for i in {1..5}; do
  code-swarm agents create --name "agent-$i" --role backend
done

Export Task Results

# Export to JSON
code-swarm tasks list --format json > tasks.json

# Export to CSV
code-swarm tasks list --format csv > tasks.csv

Debug Mode

# Enable debug logging
DEBUG=1 code-swarm api

# Verbose output
code-swarm --verbose agents list

Shell Completion

# Bash
eval "$(code-swarm --bash-complete)"

# Zsh
eval "$(code-swarm --zsh-complete)"

Help

# Global help
code-swarm --help

# Command-specific help
code-swarm agents --help
code-swarm tasks --help
code-swarm deploy --help

Last updated: 2026-05-03