Agent Chat

The Agent Chat endpoint provides an AI assistant that can autonomously use tools to search the knowledge base, recall past conversations, and store facts for future use.

Endpoint

POST https://api.solidrust.ai/v1/agent/chat

Request Parameters

Parameter	Type	Required	Default	Description
message	string	✅ Yes	—	The user’s message
system_prompt	string	No	—	Custom system prompt to guide behavior
history	object[]	No	—	Previous conversation turns
temperature	number	No	0.7	Sampling temperature (0.0–2.0)
max_tokens	integer	No	1024	Maximum response tokens

History Format

Each history entry has role (user or assistant) and content:

{
  "history": [
    {"role": "user", "content": "What is semantic search?"},
    {"role": "assistant", "content": "Semantic search finds documents by meaning..."}
  ]
}

Example Request

curl

curl -X POST "https://api.solidrust.ai/v1/agent/chat" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "message": "How do I implement RAG with Python?",
    "temperature": 0.3,
    "max_tokens": 1024
  }'

Python

import requests

API_KEY = "YOUR_API_KEY"
BASE_URL = "https://api.solidrust.ai"

def agent_chat(
    message: str,
    history: list[dict] | None = None,
    system_prompt: str | None = None,
    temperature: float = 0.7,
    max_tokens: int = 1024,
) -> dict:
    """Chat with the tool-enabled agent."""
    payload = {
        "message": message,
        "temperature": temperature,
        "max_tokens": max_tokens,
    }
    if history:
        payload["history"] = history
    if system_prompt:
        payload["system_prompt"] = system_prompt

    response = requests.post(
        f"{BASE_URL}/v1/agent/chat",
        headers={
            "Authorization": f"Bearer {API_KEY}",
            "Content-Type": "application/json",
        },
        json=payload,
    )
    return response.json()

result = agent_chat("How do I implement RAG with Python?")

if result["success"]:
    print(f"Agent: {result['response']}")

    if result.get("tool_calls"):
        print("\nTools used:")
        for call in result["tool_calls"]:
            print(f"  - {call['tool']}")
else:
    print(f"Error: {result.get('error')}")

JavaScript

async function agentChat(message, options = {}) {
  const payload = {
    message,
    temperature: options.temperature ?? 0.7,
    max_tokens: options.maxTokens ?? 1024,
  };
  if (options.history) payload.history = options.history;
  if (options.systemPrompt) payload.system_prompt = options.systemPrompt;

  const res = await fetch(`${BASE_URL}/v1/agent/chat`, {
    method: 'POST',
    headers: { Authorization: `Bearer ${API_KEY}`, 'Content-Type': 'application/json' },
    body: JSON.stringify(payload),
  });
  return res.json();
}

Example Response

{
  "success": true,
  "response": "To implement RAG with Python, you can use the SolidRusT AI API. First, search the knowledge base with hybrid search, then pass the results as context to the chat completions endpoint...",
  "tool_calls": [
    {
      "tool": "HybridSearchTool",
      "arguments": {"query": "Python RAG implementation", "limit": 5},
      "result": {"total": 3, "results": [...]}
    },
    {
      "tool": "KnowledgeGraphTool",
      "arguments": {"entity": "RAG", "depth": 1},
      "result": {"entities": [...], "relationships": [...]}
    }
  ]
}

Multi-Turn Conversations

Maintain context by passing previous turns:

conversation = []

def chat_with_history(message: str, system_prompt: str = None) -> str:
    result = agent_chat(
        message=message,
        history=conversation,
        system_prompt=system_prompt,
    )

    if result["success"]:
        conversation.append({"role": "user", "content": message})
        conversation.append({"role": "assistant", "content": result["response"]})
        return result["response"]

    return f"Error: {result.get('error')}"

# Multi-turn
print(chat_with_history("What is semantic search?"))
print(chat_with_history("How is it different from keyword search?"))
print(chat_with_history("Show me a code example"))

Custom System Prompts

system_prompt = """
You are a knowledgeable API documentation assistant.
Always cite sources from the knowledge base.
Provide code examples in Python when possible.
Keep responses concise and developer-friendly.
If you don't know, say so honestly.
"""

result = agent_chat(
    message="How do I authenticate API requests?",
    system_prompt=system_prompt,
    temperature=0.3,
)

Available Tools

The agent has access to these built-in tools:

Tool	Description
SemanticSearchTool	Vector similarity search across the knowledge base
KnowledgeGraphTool	Entity relationship queries in Neo4j
HybridSearchTool	Combined semantic + graph search with entity extraction
MemorySearchTool	Search companion memory (vector + keyword + graph + KB)
MemoryStoreTool	Store important facts for future recall
MemoryStatsTool	Check memory system health and population

See Agent Tools for complete tool schemas and parameters.

Agent Health

GET /v1/agent/health

curl "https://api.solidrust.ai/v1/agent/health" \
  -H "Authorization: Bearer YOUR_API_KEY"

Status	Meaning
healthy	All providers reachable
degraded	One provider down, fallback active
unhealthy	No providers available

Error Responses

Status	Code	Description
400	invalid_request	Missing or invalid parameters
401	INVALID_API_KEY	Missing or invalid API key
429	RATE_LIMIT_EXCEEDED	Too many requests
503	service_unavailable	No LLM provider available

Agent vs Direct Endpoints

Approach	Best For	Control
Agent Chat	Users who want automatic tool selection	Agent decides which search strategy to use
Direct RAG	Developers building custom pipelines	You control search parameters and strategy

Related

• Agent Tools — Complete tool schemas and how to use them
• Chat Completions — Direct LLM chat without tools
• RAG Applications Guide — Building RAG pipelines manually