spres-ai 1.0.0

Official Spres-Ai-Agent SDK for Python

Brainbox Python SDK

Python client library for Brainbox AI Backend.

Installation

pip install requests

Quick Start

from brainbox_sdk import BrainboxPythonSDK

sdk = BrainboxPythonSDK(
    api_url="http://localhost:8000",
    api_key="your-api-key",
    tenant_id="company-1"
)

# Ingest data
result = sdk.ingest(
    source_type="logs",
    content="ERROR: Database connection failed",
    file_path="/var/log/app.log"
)
print(f"Task ID: {result['task_id']}")

# Chat
response = sdk.chat("What happened to the database?")
print(f"Answer: {response['response']}")

Methods

ingest(source_type, content, file_path=None, metadata=None)

Send data to Brainbox for processing.

Parameters:

• source_type (str): Type of data (logs, codebase, json, csv, docker_logs, nginx_logs, postgres_logs)
• content (str): The actual content to ingest
• file_path (str, optional): File path for reference
• metadata (dict, optional): Additional metadata

Returns: Dictionary with task_id and status

chat(question, session_id=None)

Ask a question to the AI.

Parameters:

• question (str): Your question
• session_id (str, optional): Chat session ID

Returns: Dictionary with response, reasoning, and search_results

get_ingest_status(task_id)

Check ingestion task status.

Parameters:

• task_id (str): Task ID from ingest response

Returns: Task status dictionary

create_chat_session(title=None)

Create a new chat session.

Parameters:

• title (str, optional): Session title

Returns: Dictionary with session_id

health_check()

Check if backend is running.

Returns: Service health status

Examples

Ingest Logs

sdk.ingest(
    source_type="logs",
    content=open("/var/log/nginx/error.log").read(),
    file_path="/var/log/nginx/error.log"
)

Ingest Code

sdk.ingest(
    source_type="codebase",
    content=open("app.py").read(),
    file_path="app.py"
)

Ingest JSON

sdk.ingest(
    source_type="json",
    content='{"error": "Failed", "code": 500}',
    file_path="config.json"
)

Chat Session

# Create session
session = sdk.create_chat_session(title="Troubleshooting")
session_id = session['session_id']

# Ask questions in the session
response = sdk.chat("What's happening?", session_id=session_id)
response = sdk.chat("How do I fix it?", session_id=session_id)

Error Handling

try:
    response = sdk.chat("Question")
except Exception as e:
    print(f"Error: {e}")

Source Types

Type	Use Case
logs	Generic log files
nginx_logs	Nginx access/error logs
docker_logs	Docker container logs
postgres_logs	PostgreSQL logs
codebase	Source code
json	JSON formatted data
csv	CSV formatted data

Function Locator (Built-in)

Find functions in your codebase without needing a separate tool.

Find a Specific Function

sdk = BrainboxPythonSDK("http://localhost:8000", "api-key", "tenant-1")

# Find login function
login_funcs = sdk.find_function("login", directory="./src")
for func in login_funcs:
    print(f"Found: {func.name}")
    print(f"Location: {func.file_path}:{func.line_number}")
    print(f"Signature: {func.signature}")
    print(f"Parameters: {func.parameters}")

Find All Functions

# Get all functions in codebase
all_funcs = sdk.find_all_functions(directory="./src")
print(f"Total functions: {len(all_funcs)}")

Find Functions in Specific File

# Get all functions in a file
auth_funcs = sdk.find_function_by_file("./src/auth.py")
for func in auth_funcs:
    print(f"  {func.name} at line {func.line_number}")

Find Async Functions

# Find all async/await functions
async_funcs = sdk.find_async_functions(directory="./src")
print(f"Found {len(async_funcs)} async functions")

for func in async_funcs:
    print(f"  {func.name} (async) in {func.file_path}")

Function Info Object

func = sdk.find_function("login")[0]

# Access function details
print(func.name)           # "login"
print(func.file_path)      # "/app/auth.py"
print(func.line_number)    # 45
print(func.signature)      # "def login(username, password)"
print(func.parameters)     # ["username", "password"]
print(func.is_async)       # False
print(func.language)       # "python"
print(func.class_name)     # None (or class name if method)

# Convert to dict
func_dict = func.to_dict()

Supported Languages

• Python (.py files)
• JavaScript (.js files)
• TypeScript/React (.ts, .tsx files)