Cost Tracking for AI Agents

Your Agent Spent $400.
On What?

OneShot runs the action, signs a receipt for every call, and shows you where the money went. One SDK.

$npm install@oneshot-agent/sdk

Capabilities

37 tools. One call.

Every action your agent needs: messaging, research, web, commerce, build. Each one metered, authenticated, and audited.

The OneShot SDK

Email
agent.email.send({
  to: '[email protected]',
  subject: 'Update'
})
Voice Call
agent.voice.call({
  to: '+15551234567',
  script: 'intro.md'
})
SMS
agent.sms.send({
  to: '+15551234567',
  body: 'Hello!'
})
Email Inbox
agent.inbox.list({
  domain: 'myco.com',
  limit: 10
})
Read Email
agent.inbox.get({
  id: 'msg_01HX'
})
SMS Inbox
agent.sms.inbox.list({
  phone: '+15551234567',
  limit: 20
})

How It Works

See it. Control it. Prove it.

Every action your agent takes is captured the moment it happens — what it did, what it cost, whether it worked. The SDK handles the action and shows where every dollar went.

Audit Trail
Live
14:32:04
research.deep · $0.40
0xe4f2...8b23 · worker
Every Action. Accounted For.

Every Dollar Has a Receipt.
Here's One.

Each action produces a signed receipt that any machine can verify. You get a paper trail, not a trust assumption.

receipt.json
{
action"commerce.buy"
agent"agent_0x7f3a..."
amount"$42.99"
status"✓ verified"
category"commerce"
signature"0x3a9f...c2d1"
}

You See Everything

What your agent did, what it cost, whether it worked. A receipt for each action, not a log file you hope is complete.

Spend is Categorized

Inference, data, compute, commerce. The SDK tags each dollar automatically so you don't have to.

Signed by OneShot

Receipts are tamper-proof. You can audit any agent without trusting its self-reported claims.

One Import. Any Framework.

Drop OneShot into the agent stack you already use. Your receipts follow you.

LangChainlive
Virtuals GAME SDKlive
CrewAI
AutoGen
OpenAI Agents SDK
# Inside your LangChain agent
from langchain_oneshot import OneShotToolkit
toolkit = OneShotToolkit.from_private_key('0x...')
tools = toolkit.get_tools()
# payments, email, voice, research — ready

LangChain and Virtuals GAME SDK today, via the Python package. CrewAI, AutoGen, and OpenAI Agents SDK next.

Your agent runs 24/7.
Your budget shouldn't disappear.

Your agent sends 200 emails, researches 50 leads, and makes 3 voice calls before you wake up. Did it stay on budget? Did the calls convert? Which emails bounced?

You built the agent. OneShot handles the action and the accounting.

01

Receipt for every call

On-chain settlement via x402. Cryptographic proof your agent paid and delivered.

02

Budget caps that bite

Per-goal caps, daily limits, 80% alerts. Your agent stops before the budget runs out.

03

Spend analytics, live

Breakdown by tool, by day, by agent. Not a monthly CSV. A live API.

Built for teams shipping agents