langchain-veroq

LangChain tools for the VEROQ Intelligence API -- verified intelligence with confidence scores, bias ratings, and source analysis.

Installation

pip install langchain-veroq

Quick Start

from langchain_veroq import VeroqAskTool, VeroqVerifyTool

tools = [VeroqAskTool(), VeroqVerifyTool()]
# Use with any LangChain agent

Two tools cover 90% of use cases:

• VeroqAskTool -- ask any financial question in natural language
• VeroqVerifyTool -- fact-check any claim against verified intelligence

Full Agent Example

from langchain_openai import ChatOpenAI
from langchain.agents import AgentExecutor, create_openai_tools_agent
from langchain_core.prompts import ChatPromptTemplate
from langchain_veroq import VeroqAskTool, VeroqVerifyTool, VeroqSearchTool

tools = [
    VeroqAskTool(api_key="your-api-key"),
    VeroqVerifyTool(api_key="your-api-key"),
    VeroqSearchTool(api_key="your-api-key"),
]

prompt = ChatPromptTemplate.from_messages([
    ("system", "You are a financial research assistant with access to verified intelligence."),
    ("human", "{input}"),
    ("placeholder", "{agent_scratchpad}"),
])

llm = ChatOpenAI(model="gpt-4o")
agent = create_openai_tools_agent(llm, tools, prompt)
executor = AgentExecutor(agent=agent, tools=tools)

# Ask anything
result = executor.invoke({"input": "How is NVDA doing?"})
print(result["output"])

# Verify a claim
result = executor.invoke({"input": "Is it true that NVIDIA beat Q4 earnings?"})
print(result["output"])

Environment Variables

The tools accept api_key in the constructor. If omitted, the SDK checks these environment variables in order:

1. VEROQ_API_KEY
2. POLARIS_API_KEY

RAG with VeroqRetriever

from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.runnables import RunnablePassthrough
from langchain_core.output_parsers import StrOutputParser
from langchain_veroq import VeroqRetriever

retriever = VeroqRetriever(api_key="your-api-key", category="ai_ml", limit=5)

chain = (
    {"context": retriever, "question": RunnablePassthrough()}
    | ChatPromptTemplate.from_template(
        "Answer based on these verified briefs:\n\n{context}\n\nQuestion: {question}"
    )
    | ChatOpenAI(model="gpt-4o")
    | StrOutputParser()
)

print(chain.invoke("Latest developments in AI?"))

Available Tools

Backward Compatibility

This package also exports all tools under their original Polaris* names for backward compatibility. Both VeroqSearchTool and PolarisSearchTool work identically.

Documentation

Full API docs at veroq.ai/docs