Skip to content

Latest commit

 

History

History
1392 lines (1146 loc) · 26.2 KB

API.md

File metadata and controls

1392 lines (1146 loc) · 26.2 KB

AgentCraft API Reference

Complete API documentation for the AgentCraft multi-agent AI system.

Base URL

Development: http://localhost:8000
Production: https://your-domain.com

Table of Contents

  1. Authentication
  2. Agent Management
  3. Query Processing
  4. Knowledge Management
  5. Conversation Management
  6. Efficiency & Optimization
  7. WebSocket API
  8. Error Handling

Authentication

Currently, the API does not require authentication for development. For production deployment, implement API key-based or JWT authentication.

Future Authentication Header

Authorization: Bearer <your-api-key>

Agent Management

List All Agents

Get all active agents in the system.

Endpoint: GET /api/agents/list

Response:

{
  "success": true,
  "agents": {
    "agent-uuid-1": {
      "id": "uuid",
      "name": "Technical Integration Specialist",
      "role": "API and webhook integration expert",
      "domain": "technical",
      "avatar": "⚙️",
      "color": "blue",
      "keywords": ["webhook", "api", "integration"],
      "specialization_score": 0.95,
      "collaboration_rating": 0.88,
      "is_active": true
    }
  },
  "total_agents": 22
}

Create Agent

Create a new specialized agent.

Endpoint: POST /api/agents/create

Request Body:

{
  "name": "Custom Security Analyst",
  "role": "Advanced cybersecurity analysis and threat detection",
  "domain": "security",
  "backstory": "Expert cybersecurity analyst with years of experience",
  "goal": "Identify security threats and provide actionable recommendations",
  "keywords": ["security", "vulnerability", "threat", "malware"],
  "avatar": "🔒",
  "color": "red",
  "llm_config": {
    "model": "claude-3-5-sonnet-20241022",
    "temperature": 0.1
  },
  "tools": ["security_scanner", "vulnerability_detector"],
  "specialization_score": 0.9
}

Response:

{
  "success": true,
  "agent_id": "new-agent-uuid",
  "message": "Agent created successfully"
}

Update Agent

Update an existing agent's configuration.

Endpoint: PUT /api/agents/{agent_id}

Request Body:

{
  "role": "Enhanced cybersecurity analysis with AI-powered threat detection",
  "keywords": ["security", "ai", "machine-learning", "threat-detection"],
  "specialization_score": 0.95,
  "llm_config": {
    "model": "gpt-4",
    "temperature": 0.05
  }
}

Response:

{
  "success": true,
  "agent_id": "agent-uuid",
  "message": "Agent updated successfully"
}

Delete Agent

Deactivate an agent (soft delete).

Endpoint: DELETE /api/agents/{agent_id}

Response:

{
  "success": true,
  "agent_id": "agent-uuid",
  "message": "Agent deactivated successfully"
}

Get Agent Metrics

Get performance metrics for a specific agent.

Endpoint: GET /api/agents/{agent_id}/metrics

Response:

{
  "success": true,
  "agent_id": "agent-uuid",
  "metrics": {
    "total_interactions": 450,
    "avg_quality": 0.89,
    "avg_response_time_ms": 1850,
    "avg_user_rating": 4.3,
    "success_rate": 0.97,
    "avg_cost": 0.05
  }
}

Hot Reload Agent

Reload a specific agent's configuration without restarting.

Endpoint: POST /api/agents/{agent_id}/reload

Response:

{
  "success": true,
  "agent_id": "agent-uuid",
  "message": "Agent reloaded successfully"
}

Get Agent System Status

Get overall agent system status.

Endpoint: GET /api/agents/status

Response:

{
  "success": true,
  "total_agents": 22,
  "cache_stats": {
    "total_agents": 22,
    "cache_hits": 450,
    "cache_misses": 12
  },
  "agents_by_domain": {
    "technical": [
      {
        "name": "Technical Integration Specialist",
        "id": "uuid",
        "specialization_score": 0.95,
        "collaboration_rating": 0.88
      }
    ],
    "business": [...],
    "analysis": [...]
  },
  "database_connected": true,
  "memory_cached": true
}

Refresh Agent Cache

Manually refresh the agent cache.

Endpoint: POST /api/agents/cache/refresh

Response:

{
  "success": true,
  "message": "Agent cache refreshed",
  "agents_loaded": 22
}

Query Processing

Process Single Query

Process a query with automatic agent routing.

Endpoint: POST /api/query

Request Body:

{
  "query": "How do I fix webhook signature validation errors?",
  "context": {
    "user_id": "user-123",
    "session_id": "session-456"
  }
}

Response:

{
  "success": true,
  "response": "Detailed technical response...",
  "routing_info": {
    "selected_agent": "Technical Integration Specialist",
    "confidence": 0.85,
    "routing_time": 0.05
  },
  "citations": [
    {
      "source": "Zapier Official Documentation",
      "url": "https://zapier.com/help/webhooks",
      "relevance": 0.92
    }
  ],
  "session_id": "session-uuid",
  "response_time_ms": 1500,
  "tokens_used": 2500,
  "cost": 0.05
}

Multi-Agent Query Processing

Process query with multiple specialized agents.

Endpoint: POST /api/multi-agent/query

Request Body:

{
  "query": "Analyze the security implications of our webhook implementation",
  "session_id": "session-uuid",
  "context": {
    "user_id": "user-123",
    "priority": "high"
  }
}

Response:

{
  "success": true,
  "response": "Combined response from multiple agents...",
  "agents_used": [
    "Technical Integration Specialist",
    "Security Specialist",
    "Database Expert"
  ],
  "collaboration_log": [
    {
      "timestamp": "2024-01-20T10:30:00Z",
      "agent": "Technical Integration Specialist",
      "action": "initial_analysis",
      "output": "..."
    },
    {
      "timestamp": "2024-01-20T10:30:15Z",
      "agent": "Security Specialist",
      "action": "security_review",
      "output": "..."
    }
  ],
  "citations": [...],
  "performance": {
    "response_time_ms": 2500,
    "tokens_used": 4200,
    "cost": 0.08,
    "quality_score": 0.93
  },
  "session_id": "session-uuid",
  "database_session_id": "db-session-uuid"
}

Knowledge Management

Search Knowledge Base

Search through the knowledge base using semantic search.

Endpoint: GET /api/knowledge/search

Query Parameters:

  • q (required): Search query
  • limit (optional): Number of results (default: 10, max: 50)
  • company (optional): Filter by company name

Example:

GET /api/knowledge/search?q=webhook%20authentication&limit=5

Response:

{
  "success": true,
  "results": [
    {
      "id": "doc-uuid",
      "title": "Webhook Authentication Best Practices",
      "content": "Detailed content...",
      "relevance_score": 0.92,
      "source": "qdrant",
      "url": "https://example.com/docs/webhooks",
      "category": "documentation",
      "created_at": "2024-01-20T10:00:00Z"
    }
  ],
  "total_results": 5,
  "query": "webhook authentication"
}

Add Crawl URL

Add a URL to be crawled for knowledge base.

Endpoint: POST /api/knowledge/urls

Request Body:

{
  "company_name": "Zapier",
  "url": "https://zapier.com/help/webhooks"
}

Response:

{
  "success": true,
  "url_id": "url-uuid",
  "company_name": "Zapier",
  "url": "https://zapier.com/help/webhooks",
  "message": "URL added to crawl queue"
}

Remove Crawl URL

Remove a URL from the crawl list.

Endpoint: DELETE /api/knowledge/urls

Request Body:

{
  "company_name": "Zapier",
  "url": "https://zapier.com/help/webhooks"
}

Response:

{
  "success": true,
  "message": "URL removed from crawl list"
}

Get Company URLs

Get all crawl URLs for a specific company.

Endpoint: GET /api/knowledge/companies/{company_name}/urls

Response:

{
  "success": true,
  "company_name": "Zapier",
  "urls": [
    {
      "id": "url-uuid",
      "url": "https://zapier.com/help/webhooks",
      "is_active": true,
      "created_at": "2024-01-20T10:00:00Z",
      "last_crawled": "2024-01-20T15:30:00Z"
    }
  ],
  "total_urls": 10
}

Trigger Knowledge Crawl

Trigger Firecrawl to scrape and index content.

Endpoint: POST /api/knowledge/crawl

Request Body:

{
  "company_name": "Zapier",
  "max_pages": 50,
  "force_refresh": false
}

Response:

{
  "success": true,
  "job_id": "crawl-job-uuid",
  "status": "started",
  "company_name": "Zapier",
  "max_pages": 50,
  "estimated_completion": "2024-01-20T16:00:00Z"
}

Check Crawl Job Status

Check the status of a crawl job.

Endpoint: GET /api/knowledge/crawl/{job_id}

Response:

{
  "success": true,
  "job_id": "crawl-job-uuid",
  "status": "completed",
  "company_name": "Zapier",
  "pages_crawled": 42,
  "documents_indexed": 38,
  "started_at": "2024-01-20T15:00:00Z",
  "completed_at": "2024-01-20T15:45:00Z",
  "errors": []
}

Get All Companies

List all companies in the knowledge base.

Endpoint: GET /api/knowledge/companies

Response:

{
  "success": true,
  "companies": [
    {
      "id": "company-uuid",
      "name": "Zapier",
      "domain": "zapier.com",
      "description": "Automation platform",
      "url_count": 10,
      "is_active": true,
      "created_at": "2024-01-15T10:00:00Z"
    }
  ],
  "total_companies": 5
}

Set Current Company

Set the active company for knowledge retrieval.

Endpoint: POST /api/knowledge/companies/current

Request Body:

{
  "company_name": "Zapier"
}

Response:

{
  "success": true,
  "current_company": "Zapier",
  "message": "Current company updated"
}

Conversation Management

Get Conversation History

Retrieve conversation history for a session.

Endpoint: GET /api/conversation/{session_id}

Response:

{
  "success": true,
  "session_id": "session-uuid",
  "messages": [
    {
      "role": "user",
      "content": "How do I validate webhook signatures?",
      "timestamp": "2024-01-20T10:30:00Z"
    },
    {
      "role": "assistant",
      "content": "Here's how to validate webhook signatures...",
      "timestamp": "2024-01-20T10:30:02Z",
      "agent": "Technical Integration Specialist",
      "citations": [...]
    }
  ],
  "agents_used": ["Technical Integration Specialist"],
  "created_at": "2024-01-20T10:30:00Z",
  "message_count": 6
}

List Active Conversations

Get all active conversation sessions.

Endpoint: GET /api/conversations

Query Parameters:

  • limit (optional): Number of conversations (default: 50)
  • offset (optional): Pagination offset (default: 0)

Response:

{
  "success": true,
  "sessions": [
    {
      "session_id": "session-uuid",
      "query": "First message...",
      "created_at": "2024-01-20T10:00:00Z",
      "last_activity": "2024-01-20T10:30:00Z",
      "message_count": 8,
      "agents_used": ["Technical Integration Specialist"]
    }
  ],
  "total_sessions": 156,
  "limit": 50,
  "offset": 0
}

Submit User Feedback

Submit feedback for a conversation session.

Endpoint: POST /api/agents/feedback

Request Body:

{
  "session_id": "session-uuid",
  "rating": 5,
  "comment": "Excellent technical support, solved my issue quickly"
}

Response:

{
  "success": true,
  "session_id": "session-uuid",
  "message": "Feedback recorded successfully",
  "learning_insight_generated": true
}

Efficiency & Optimization

Bulk Chat Processing

Process multiple messages efficiently in parallel.

Endpoint: POST /api/efficiency/bulk-chat

Request Body:

{
  "messages": [
    "How do I set up webhooks?",
    "What's the billing structure?",
    "Security best practices?"
  ],
  "agent_type": "multi-agent",
  "context": {
    "user_id": "user-123"
  },
  "priority": "normal"
}

Response:

{
  "success": true,
  "processed_count": 3,
  "results": [
    {
      "index": 0,
      "message": "How do I set up webhooks?",
      "result": {
        "success": true,
        "response": "...",
        "database_session_id": "session-uuid-1"
      }
    }
  ],
  "session_ids": ["session-uuid-1", "session-uuid-2", "session-uuid-3"],
  "bulk_processing": true
}

Get Chat Templates

Get predefined templates for common queries.

Endpoint: GET /api/efficiency/templates

Response:

{
  "success": true,
  "templates": [
    {
      "name": "webhook_troubleshooting",
      "template": "I'm having issues with webhook {webhook_url}. The error is: {error_message}. Can you help diagnose the problem?",
      "variables": ["webhook_url", "error_message"],
      "category": "technical",
      "description": "Template for webhook integration issues"
    }
  ],
  "total_count": 5
}

Generate from Template

Generate and process a message from a template.

Endpoint: POST /api/efficiency/templates/{template_name}/generate

Request Body:

{
  "webhook_url": "https://api.example.com/webhook",
  "error_message": "401 Unauthorized"
}

Response:

{
  "success": true,
  "template_name": "webhook_troubleshooting",
  "generated_message": "I'm having issues with webhook https://api.example.com/webhook. The error is: 401 Unauthorized. Can you help diagnose the problem?",
  "result": {
    "success": true,
    "response": "..."
  }
}

Get Agent Presets

Get predefined agent combinations for specific use cases.

Endpoint: GET /api/efficiency/agent-presets

Response:

{
  "success": true,
  "presets": [
    {
      "name": "technical_support",
      "description": "Best agents for technical troubleshooting",
      "agent_ids": ["agent-uuid-1", "agent-uuid-2"],
      "use_case": "API issues, webhook problems",
      "priority_order": [0, 1, 2]
    }
  ],
  "total_presets": 3
}

Get Performance Insights

Get performance insights and optimization recommendations.

Endpoint: GET /api/efficiency/performance-insights

Query Parameters:

  • days (optional): Time period in days (default: 7, max: 90)
  • limit (optional): Number of insights (default: 10, max: 50)

Response:

{
  "success": true,
  "performance_score": 92.0,
  "time_period_days": 7,
  "insights": {
    "query_patterns": [
      {
        "pattern_description": "Webhook troubleshooting",
        "frequency": 45,
        "avg_satisfaction": 4.5,
        "optimal_agents": ["Technical Integration Specialist"]
      }
    ],
    "learning_insights": [
      {
        "title": "High user satisfaction on security queries",
        "type": "user_satisfaction",
        "confidence": 0.9,
        "data_points": 23
      }
    ],
    "recommendations": [
      {
        "type": "optimization",
        "priority": "medium",
        "title": "Optimize Common Query Pattern",
        "description": "..."
      }
    ]
  },
  "system_health": {...},
  "timestamp": "2024-01-20T10:30:00Z"
}

Optimize System

Trigger system optimization tasks.

Endpoint: POST /api/efficiency/optimize

Request Body:

{
  "optimization_type": "cache",
  "parameters": {
    "force_refresh": true
  }
}

Optimization Types:

  • cache: Refresh agent cache
  • performance: Performance optimization
  • learning: Generate learning insights
  • agents: Agent optimization

Response:

{
  "success": true,
  "optimization_type": "cache",
  "results": {
    "cache_refresh": {
      "success": true,
      "agents_loaded": 22
    }
  },
  "timestamp": "2024-01-20T10:30:00Z"
}

Get Usage Statistics

Get comprehensive usage statistics.

Endpoint: GET /api/efficiency/usage-statistics

Query Parameters:

  • days (optional): Time period (default: 7, max: 90)

Response:

{
  "success": true,
  "statistics": {
    "time_period": {
      "start_date": "2024-01-13T00:00:00Z",
      "end_date": "2024-01-20T00:00:00Z",
      "days": 7
    },
    "conversation_stats": {
      "total_conversations": 156,
      "successful_conversations": 142,
      "success_rate": 91.0,
      "avg_response_time": "2.3s",
      "avg_user_satisfaction": 4.2
    },
    "agent_usage": {
      "most_used_agent": "Technical Integration Specialist",
      "avg_agents_per_query": 2.1,
      "collaboration_rate": 68.0
    },
    "performance_metrics": {
      "cache_hit_rate": 89.5,
      "database_response_time": "45ms",
      "memory_efficiency": 92.3
    },
    "learning_metrics": {
      "insights_generated": 23,
      "insights_implemented": 18,
      "improvement_rate": 78.3
    }
  },
  "generated_at": "2024-01-20T10:30:00Z"
}

Comprehensive Health Check

Get comprehensive system health status.

Endpoint: GET /api/efficiency/health-check

Response:

{
  "overall_status": "healthy",
  "timestamp": "2024-01-20T10:30:00Z",
  "checks": {
    "database": {
      "status": "healthy",
      "pool_size": 5
    },
    "agent_cache": {
      "status": "healthy",
      "stats": {
        "total_agents": 22,
        "cache_hits": 450,
        "cache_misses": 12
      }
    },
    "websockets": {
      "status": "healthy",
      "active_sessions": 8
    },
    "crewai": {
      "status": "healthy",
      "patched": true
    }
  }
}

WebSocket API

Agent Tracking WebSocket

Real-time agent status updates.

Endpoint: ws://localhost:8000/ws/agent-tracking/{client_id}

Connection:

const ws = new WebSocket('ws://localhost:8000/ws/agent-tracking/client-123');

ws.onopen = () => {
  console.log('WebSocket connected');
};

ws.onmessage = (event) => {
  const message = JSON.parse(event.data);
  console.log('Received:', message);
};

Message Types Received:

1. Active Sessions

{
  "type": "active_sessions",
  "timestamp": 1642678200.123,
  "data": {
    "total_sessions": 3,
    "sessions": [...]
  }
}

2. Agent Status Update

{
  "type": "agent_status",
  "agent_id": "agent-uuid",
  "agent_name": "Technical Integration Specialist",
  "status": "PROCESSING",
  "progress": 65,
  "message": "Analyzing webhook signature validation...",
  "timestamp": "2024-01-20T10:30:45.123Z",
  "metadata": {
    "task_id": "task-uuid",
    "estimated_completion": 30
  }
}

3. Agent Collaboration

{
  "type": "agent_collaboration",
  "primary_agent": "Technical Integration Specialist",
  "secondary_agent": "Security Specialist",
  "collaboration_type": "handoff",
  "reason": "Requires security expertise",
  "timestamp": "2024-01-20T10:31:15.456Z"
}

4. Ping/Pong

{
  "type": "ping",
  "timestamp": 1642678200.123
}

Client Messages to Send:

1. Get Session Logs

{
  "type": "get_session_logs",
  "session_id": "session-uuid"
}

2. Get Session State

{
  "type": "get_session_state",
  "session_id": "session-uuid"
}

3. Start Log Streaming

{
  "type": "start_log_streaming",
  "session_id": "session-uuid"
}

4. Stop Log Streaming

{
  "type": "stop_log_streaming"
}

5. Pong (response to ping)

{
  "type": "pong"
}

WebSocket Stats

Get WebSocket connection statistics (REST endpoint).

Endpoint: GET /api/ws/stats

Response:

{
  "success": true,
  "stats": {
    "active_connections": 5,
    "total_messages_sent": 1250,
    "connections": [
      {
        "client_id": "client-123",
        "connected_at": 1642678000.0,
        "message_count": 45
      }
    ]
  },
  "realtime_sessions": {
    "total_sessions": 3,
    "active_agents": 8
  }
}

Get Session Logs (REST)

Get execution logs for a session (REST endpoint).

Endpoint: GET /api/ws/sessions/{session_id}/logs

Response:

{
  "success": true,
  "session_id": "session-uuid",
  "logs": [
    {
      "timestamp": "2024-01-20T10:30:00Z",
      "level": "info",
      "agent": "Technical Integration Specialist",
      "message": "Starting webhook analysis",
      "details": {}
    }
  ],
  "total_logs": 45
}

Get Session State (REST)

Get current session state (REST endpoint).

Endpoint: GET /api/ws/sessions/{session_id}/state

Response:

{
  "success": true,
  "session_id": "session-uuid",
  "state": {
    "session_id": "session-uuid",
    "query": "How do I fix webhook errors?",
    "total_agents": 3,
    "current_phase": "processing",
    "overall_progress": 65,
    "started_at": "2024-01-20T10:30:00Z",
    "estimated_completion": 30,
    "active_agents": [
      {
        "agent_id": "agent-uuid",
        "agent_name": "Technical Integration Specialist",
        "status": "PROCESSING",
        "current_task": "Analyzing webhook configuration",
        "progress": 75,
        "details": {},
        "started_at": "2024-01-20T10:30:05Z",
        "updated_at": "2024-01-20T10:30:20Z"
      }
    ],
    "completed_agents": [],
    "crew_output": [...]
  }
}

Error Handling

Error Response Format

All errors follow this format:

{
  "detail": "Error message description",
  "status_code": 400,
  "error_type": "validation_error",
  "timestamp": "2024-01-20T10:30:00Z"
}

HTTP Status Codes

Code Meaning Description
200 OK Request succeeded
201 Created Resource created successfully
400 Bad Request Invalid request parameters
401 Unauthorized Authentication required
403 Forbidden Insufficient permissions
404 Not Found Resource not found
422 Unprocessable Entity Validation error
429 Too Many Requests Rate limit exceeded
500 Internal Server Error Server error
503 Service Unavailable Service temporarily unavailable

Common Error Scenarios

1. Invalid Agent ID

{
  "detail": "Invalid agent ID format",
  "status_code": 400
}

2. Agent Not Found

{
  "detail": "Agent not found",
  "status_code": 404
}

3. Session Not Found

{
  "detail": "Session not found",
  "status_code": 404
}

4. Database Connection Error

{
  "detail": "Database connection failed",
  "status_code": 503
}

5. Validation Error

{
  "detail": "Validation error: name is required",
  "status_code": 422,
  "validation_errors": [
    {
      "field": "name",
      "message": "Field required"
    }
  ]
}

Rate Limiting

Currently not implemented. For production:

Rate Limits:
- 100 requests per minute per IP
- 1000 requests per hour per API key

Rate Limit Headers:

X-RateLimit-Limit: 100
X-RateLimit-Remaining: 95
X-RateLimit-Reset: 1642678260

Examples

Complete Workflow Example

// 1. Create a custom agent
const createResponse = await fetch('http://localhost:8000/api/agents/create', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    name: "Custom Security Analyst",
    role: "Cybersecurity expert",
    domain: "security",
    keywords: ["security", "vulnerability"],
    llm_config: {
      model: "claude-3-5-sonnet-20241022",
      temperature: 0.1
    }
  })
});

const { agent_id } = await createResponse.json();

// 2. Process a query
const queryResponse = await fetch('http://localhost:8000/api/multi-agent/query', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    query: "Analyze security vulnerabilities in our webhook system",
    context: { user_id: "user-123" }
  })
});

const result = await queryResponse.json();
console.log('Response:', result.response);
console.log('Agents used:', result.agents_used);

// 3. Submit feedback
await fetch('http://localhost:8000/api/agents/feedback', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    session_id: result.session_id,
    rating: 5,
    comment: "Excellent security analysis"
  })
});

WebSocket Real-Time Tracking Example

const ws = new WebSocket('ws://localhost:8000/ws/agent-tracking/my-client-id');

ws.onopen = () => {
  console.log('Connected to agent tracking');

  // Request session state
  ws.send(JSON.stringify({
    type: 'get_session_state',
    session_id: 'session-uuid'
  }));
};

ws.onmessage = (event) => {
  const message = JSON.parse(event.data);

  switch (message.type) {
    case 'agent_status':
      console.log(`Agent ${message.agent_name}: ${message.status} - ${message.message}`);
      updateProgressBar(message.progress);
      break;

    case 'agent_collaboration':
      console.log(`Collaboration: ${message.primary_agent} → ${message.secondary_agent}`);
      break;

    case 'ping':
      ws.send(JSON.stringify({ type: 'pong' }));
      break;
  }
};

ws.onerror = (error) => {
  console.error('WebSocket error:', error);
};

ws.onclose = () => {
  console.log('WebSocket closed, reconnecting...');
  setTimeout(connectWebSocket, 3000);
};

API Versioning

Current version: v1 (implicit, no version in URL)

Future versions will use URL versioning:

/api/v2/agents/list

Interactive API Documentation

FastAPI provides interactive API documentation:

These interfaces allow you to:

  • Explore all endpoints
  • Test API calls directly
  • View request/response schemas
  • See validation rules

SDK Support

JavaScript/TypeScript

import { AgentCraftClient } from './services/api';

const client = new AgentCraftClient('http://localhost:8000');

// List agents
const agents = await client.agents.list();

// Process query
const result = await client.query.process({
  query: "How do I fix webhook errors?",
  context: {}
});

// Submit feedback
await client.feedback.submit({
  session_id: result.session_id,
  rating: 5
});

Python

from agentcraft_client import AgentCraftClient

client = AgentCraftClient("http://localhost:8000")

# List agents
agents = client.agents.list()

# Process query
result = client.query.process(
    query="How do I fix webhook errors?",
    context={}
)

# Submit feedback
client.feedback.submit(
    session_id=result["session_id"],
    rating=5
)

For more information, see: