DevToolBox

Additional Recommendations

Set up automatic code review for every pull request:

# 1. Add your GitHub forge
/git add github https://github.com --token ghp_your_token --alias gh

# 2. Create a code-review skill
/skill create pr-review

# 3. Create automation triggered on PR open
curl -X POST http://dataworker:8892/api/v1/automations \
  -H "Content-Type: application/json" \
  -d '{
    "name": "auto-review-pr",
    "trigger": {"type": "git_pr_opened", "forge": "gh", "repo": "myorg/myrepo"},
    "actions": [
      {"skill": "code-review", "inputs": {"repo": "{{trigger.repo}}", "pr": "{{trigger.pr_number}}"}},
      {"type": "comment", "target": "{{trigger.pr_url}}", "body": "## AI Review\n{{skill.output.report}}"}
    ]
  }'

# 4. Every new PR now gets automatic review!

Safely run untrusted code in isolated containers:

# Create a Python sandbox with resource limits
curl -X POST http://dataworker:8892/api/v1/sandboxes \
  -H "Content-Type: application/json" \
  -d '{
    "name": "python-dev",
    "image": "python:3.12-slim",
    "limits": {"cpu": "2", "memory": "2g", "timeout_seconds": 300},
    "network": {"mode": "restricted", "allowed_hosts": ["pypi.org"]}
  }'

# Execute code in the sandbox
curl -X POST http://dataworker:8892/api/v1/sandboxes/sbx-abc123/exec \
  -H "Content-Type: application/json" \
  -d '{
    "command": "pip install numpy && python -c \"import numpy; print(numpy.random.rand(3,3))\""
  }'

# Response:
{
  "exit_code": 0,
  "output": "[[0.123 0.456 0.789]\n [0.234 0.567 0.890]\n [0.345 0.678 0.901]]",
  "duration_ms": 2340
}

Automatically analyze and label issues:

# Analyze a GitHub issue
curl -X POST http://dataworker:8892/api/v1/triage/analyze \
  -H "Content-Type: application/json" \
  -d '{"issue": "github:myorg/myrepo#42"}'

# Response:
{
  "labels": ["bug", "security", "high-priority"],
  "priority": "P1",
  "category": "security-vulnerability",
  "suggested_assignees": ["alice", "bob"],
  "related_issues": [38, 41],
  "duplicate_of": null,
  "fix_suggestion": {
    "description": "SQL injection in user input - sanitize query parameters",
    "files": ["src/api/users.py:145", "src/api/users.py:167"],
    "confidence": 0.92,
    "code_snippet": "# Replace:\nquery = f\"SELECT * FROM users WHERE id = {user_id}\"\n# With:\nquery = \"SELECT * FROM users WHERE id = %s\"\ncursor.execute(query, (user_id,))"
  }
}

Diagnose and fix build failures automatically:

# Monitor a repository's CI/CD
/cicd add github:myorg/myrepo

# When a build fails, analyze it
curl -X POST http://dataworker:8892/api/v1/cicd/analyze \
  -H "Content-Type: application/json" \
  -d '{"repo": "github:myorg/myrepo", "run_id": "12345"}'

# Response:
{
  "status": "failed",
  "failed_job": "test",
  "failed_step": "Run pytest",
  "error_type": "test_failure",
  "error_summary": "2 tests failed in tests/test_api.py",
  "root_cause": "Missing mock for external API call in test_user_creation",
  "fix_suggestion": {
    "description": "Add mock for external service",
    "file": "tests/test_api.py",
    "line": 45,
    "patch": "@mock.patch('app.services.external_api.create_user')\ndef test_user_creation(mock_create):\n    mock_create.return_value = {'id': 1, 'status': 'created'}\n    ..."
  },
  "similar_failures": [
    {"run_id": "12340", "date": "2024-02-01", "same_root_cause": true}
  ]
}

Create a skill that combines multiple agents:

# Create a "security-audit" skill
curl -X POST http://dataworker:8892/api/v1/skills \
  -H "Content-Type: application/json" \
  -d '{
    "id": "security-audit",
    "name": "Security Audit",
    "description": "Comprehensive security review of codebase",
    "version": "1.0.0",
    "inputs": [
      {"name": "repo", "type": "git_repo", "required": true},
      {"name": "severity_threshold", "type": "string", "default": "medium"}
    ],
    "steps": [
      {
        "id": "scan",
        "agent": "explorer",
        "action": "find_files",
        "params": {"patterns": ["*.py", "*.js", "*.ts"]},
        "output": "source_files"
      },
      {
        "id": "analyze",
        "agent": "coder",
        "action": "security_scan",
        "input": "source_files",
        "params": {"checks": ["sql_injection", "xss", "secrets", "dependencies"]},
        "output": "vulnerabilities"
      },
      {
        "id": "prioritize",
        "agent": "validator",
        "action": "rank_issues",
        "input": "vulnerabilities",
        "params": {"threshold": "{{severity_threshold}}"},
        "output": "prioritized_issues"
      },
      {
        "id": "report",
        "agent": "planner",
        "action": "generate_report",
        "input": "prioritized_issues",
        "output": "security_report"
      }
    ],
    "outputs": ["security_report", "prioritized_issues"]
  }'

# Run the skill
/skill run security-audit --repo github:myorg/myrepo --severity_threshold high

Trigger Types

Dashboard URL: http://controller:8880/dashboard/integrations

Access the visual workflow editor from the Controller dashboard. No code required.

Dashboard URL: http://dataworker:8892/dashboard/devtoolbox

The dashboard provides a visual interface for managing Git forges, skills, automations, sandboxes, issue triage, and CI/CD monitoring.