NORNRMandates, approvals and evidence for autonomous agents.
Guide / Flowise
10 minutesHow to add a budget to a Flowise agent
Gate paid tool calls in a Flowise chatflow with a governed wallet, approval threshold and audit trail using NORNR in under 10 minutes.
1. Why Flowise agents need a budget
Flowise makes it fast to build LLM-powered chatflows visually. But when those flows call paid models or external services, there is no native gate between the chatflow trigger and the billing event. NORNR adds that gate as a custom tool node.
2. Install the SDK
bash
pip install agentpay3. Create the governed wallet in a custom tool
python
from agentpay import Wallet
wallet = Wallet.create(
owner="flowise-research-chatflow",
daily_limit=40,
require_approval_above=10,
base_url="https://nornr.com",
)4. Gate the paid step
python
decision = wallet.pay(
amount=7.00,
to="openai",
purpose="chatflow model inference",
)
if decision.get("status") == "approved":
return {"proceed": True}
elif decision.get("status") == "queued":
return {"proceed": False, "reason": "Queued for operator review."}
else:
return {"proceed": False, "reason": f"Blocked: {decision.get('reasons')}"}5. Decision output
output
{"status": "approved", "requiresApproval": false}
{"status": "queued", "requiresApproval": true}
{"status": "rejected", "reasons": ["daily_limit_exceeded"]}6. Common mistakes
- Placing the NORNR tool after the LLM node instead of before it.
- Using a single wallet for multiple unrelated chatflows without separate owner identifiers.
- Ignoring the queued state — route it to a notification rather than silently dropping.