langchain-pqsafe 0.1.0

LangChain integration for PQSafe AgentPay — post-quantum safe payments for AI agents

langchain-pqsafe

Post-quantum safe payments for LangChain agents, powered by PQSafe AgentPay.

Built on @pqsafe/agent-pay — see github.com/PQSafe/pqsafe

---

Install

pip install langchain-pqsafe

5-line quick start

from langchain_pqsafe import PQSafePaymentTool
from langchain.agents import create_react_agent
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(model="gpt-4o")
agent = create_react_agent(llm, tools=[PQSafePaymentTool()])

That's it. The agent can now call pqsafe_pay whenever it needs to make a payment authorized by a SpendEnvelope.

---

How it works

1. A human wallet owner issues a signed SpendEnvelope — a post-quantum (ML-DSA-65) token that specifies: agent ID, max amount, allowed recipients, currency, and validity window.
2. The LangChain agent calls pqsafe_pay with the envelope, recipient, amount, and memo.
3. PQSafe verifies the envelope server-side and routes the payment over the cheapest available rail (Airwallex, Wise, Stripe, USDC/Base, or x402).
4. The tool returns {txId, status, rail}.

---

Full example

import json
from langchain_pqsafe import PQSafePaymentTool
from langchain.agents import AgentExecutor, create_react_agent
from langchain_openai import ChatOpenAI
from langchain_core.prompts import PromptTemplate

# 1. Build the tool
payment_tool = PQSafePaymentTool()

# 2. Your signed envelope (issued by the wallet owner, stored securely)
signed_envelope = {
    "envelopeJson": '{"version":1,"issuer":"pq1...","agent":"my-agent",...}',
    "signature": "deadbeef...",
    "dsaPublicKey": "cafebabe...",
}
envelope_json = json.dumps(signed_envelope)

# 3. Create the agent
llm = ChatOpenAI(model="gpt-4o", temperature=0)
prompt = PromptTemplate.from_template(
    "You are a payment agent. Use pqsafe_pay to pay invoices.\n"
    "Envelope: {envelope_json}\n\n{agent_scratchpad}"
)
agent = create_react_agent(llm, tools=[payment_tool], prompt=prompt)
executor = AgentExecutor(agent=agent, tools=[payment_tool], verbose=True)

# 4. Run
result = executor.invoke({
    "envelope_json": envelope_json,
    "input": "Renew Perplexity Pro at perplexity.ai — $20 USD for research-agent-v1",
})
print(result["output"])

---

Tool schema

Parameter	Type	Required	Description
envelope_json	str	Yes	SignedEnvelope serialized as JSON
recipient	str	Yes	Recipient address (IBAN, crypto addr, etc.)
amount	float	Yes	Amount in envelope's currency (> 0)
memo	str	No	Human-readable reference

Returns a string: Payment successful. txId=<id> status=<status> rail=<rail>

---

Links

• Documentation: docs.pqsafe.xyz
• Website: pqsafe.xyz
• GitHub: github.com/PQSafe/pqsafe
• npm SDK: @pqsafe/agent-pay