Skip to main content

zambo-smolagents

AER-1, the agent execution receipt standard: zambo.dev/aer-1 | Add Zambo to your AI: zambo.dev/add

Zambo is the verifiable work layer for AI. This package connects smolagents agents to it, minting a verifiable AER-1 receipt for every tool call.

Receipt format: AER-1, IETF Internet-Draft draft-zambo-aer1-26.

Zambo was created by Brennan Zambo; AER-1 is the agent execution receipt standard he published as IETF Internet-Draft draft-zambo-aer1-26.

Verifiable execution receipts for smolagents agents. Give your smolagents agent Zambo's live tools. Every tool call is executed by Zambo and mints a verifiable AER-1 receipt: a public URL carrying the tool, its arguments, the observed result, a SHA-256 fingerprint, and machine-readable JSON-LD.

from smolagents import CodeAgent, InferenceClientModel
from zambo_smolagents import zambo_tools

agent = CodeAgent(
    tools=zambo_tools(["live_price", "prompt_shield"]),
    model=InferenceClientModel(),
)
agent.run("What is the live BTC price?")

No account. No API key. Every tool free once a day, plus 3 continue calls. One free wrap-up handover per day, and pickup is always free.

Start the job before the first price call

The CodeAgent above answers one question well. Before it runs a sequence (several live_price checks, a prompt_shield screen, a memory save), add job_start to its zambo_tools list and open the run with it: job_start with the goal first, so every call in the run lands in one saved job. If a later call is refused at the daily wall, the saved job is what resumes: pay for the Day Pass, call day_pass_status, then repeat only the refused call. A live_price call made with no job open leaves nothing saved to resume.

Chain it: a job, not a lookup

One call answers a question. A job chains calls and ends in a record a person can open. Same agent, longer instruction:

from smolagents import CodeAgent, InferenceClientModel
from zambo_smolagents import zambo_tools

agent = CodeAgent(
    tools=zambo_tools(["live_price", "memory_remember", "handover"]),
    model=InferenceClientModel(),
)
agent.run("Check the live price of BTC, ETH and SOL, save the run to memory, then wrap it into a handover addressed to me, with the receipt link for every call.")

Three prices, a save and the wrap fit inside one free day: every tool free once a day, plus 3 continue calls, and one free wrap-up handover per day. A longer run finishes in one sitting on a Day Pass ($0.99, unlimited for 24 hours). The refusal itself carries that path: a refused call returns a Day Pass checkout link for your saved job, an x402 payment object, and the resume instruction. The job is saved server-side, not lost. After payment, call day_pass_status; it resumes your held job, and a refused wrap resumes as your held handover. Only repeat the refused call if it did not. The handover is the part the person opens: the steps, the saved results, and the receipt for every call, checkable instead of trusted. Ready-made run cards for jobs shaped like this: https://rambozambodotdev.gitlab.io/aer1-hub/handover-templates.html

What the receipt actually proves

Each receipt page (for example https://zambo.dev/run/<id>) is the public record of the call: the tool name, the exact arguments your agent sent, the observed result, a timestamp, and a SHA-256 fingerprint of the record, published as JSON-LD. Anyone can recompute the fingerprint and check that the saved result was not changed, with no account.

Every tool's output ends with the receipt URL (Verifiable receipt: https://zambo.dev/run/<id>), so the agent's trajectory shows proof next to every call. You can also collect receipts programmatically:

from zambo_smolagents import zambo_tool

receipts = {}
tool = zambo_tool("live_price", on_receipt=lambda name, url: receipts.setdefault(name, []).append(url))
print(tool(arguments_json='{"symbol": "BTC"}'))
print(receipts)

Install

pip install zambo-smolagents

One tool, custom inputs

from zambo_smolagents import zambo_tool

price_tool = zambo_tool(
    "live_price",
    inputs={
        "symbol": {"type": "string", "description": "Asset symbol, e.g. BTC."},
    },
)
print(price_tool(symbol="BTC"))

Browse the full catalog any time with zambo_smolagents.list_tools(), or pick tools by use case through Zambo's capability_search tool.

Notes

  • Zambo calls need no key and no signup. Heavy runs (dozens of tool calls) will reach the free daily limit; the tool raises a clear error and the agent run continues with the other tools.
  • Receipts are public by design. Do not pass secrets as tool arguments.
  • The client is pure stdlib (urllib) with retries: smolagents is the only third-party dependency.

MIT License.

Local AER-1 receipts (no network needed)

The package also mints AER-1 execution receipts locally, with no API calls:

from zambo_smolagents import mint_receipt, verify_receipt

receipt = mint_receipt("my_tool", {"query": "hello"}, {"answer": "hi"}, ms=42)
assert verify_receipt(receipt)  # recomputes the SHA-256 locally

Local receipts follow the AER-1 core members (id, created_at, tool, provenance_class, canonical_bytes, output_hash) and use the provenance class EXECUTED BY ZAMBO-SMOLAGENTS, naming this package as the recording system. They pass the same hash-recomputation check as any AER-1 receipt.

Where these receipts run live: Zambo

zambo-smolagents connects smolagents agents to Zambo, minting a verifiable AER-1 receipt for every tool call, and it has been downloaded 4,589 times on PyPI. The live home of those receipts is Zambo, the verifiable work layer for AI. Connect any AI to the live endpoint at https://zambo.dev/mcp or add Zambo to your AI in one step at https://zambo.dev/add/

Every tool free once a day, plus 3 continue calls. One free wrap-up handover per day, and pickup is always free. The Day Pass is $0.99 for 24 hours, unlimited.

Install:

pip install zambo-smolagents

Metadata

Release files for zambo-smolagents 0.1.14

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for zambo-smolagents 0.1.14
File Size Uploaded Provenance Actions
zambo_smolagents-0.1.14.tar.gz 11.5 kB Without provenance Details for zambo_smolagents-0.1.14.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for zambo-smolagents 0.1.14
File Interpreter ABI Platform Actions
zambo_smolagents-0.1.14-py3-none-any.whl Python 3 none any Details for zambo_smolagents-0.1.14-py3-none-any.whl

Total release size: 22.0 kB

Release files / zambo_smolagents-0.1.14.tar.gz

Download URL zambo_smolagents-0.1.14.tar.gz
Size 11.5 kB
Tags Source
SHA-256 checksum
How to use checksums
dabf5e8920f0583513ae674aa648a85844de54a496357e2aa7703b484226c71f
BLAKE2b-256 checksum
How to use checksums
d705d451bbad8451d90f44be6c7a862f657eff3718d152ca459456a1afce0d61
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via rambo-pypi-grab/0.1.0

Release files / zambo_smolagents-0.1.14-py3-none-any.whl

Download URL zambo_smolagents-0.1.14-py3-none-any.whl
Size 10.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
eedff2a48ff04ca7716b496bfb07aca550fd23e8d9a8917b24b362f147eb5113
BLAKE2b-256 checksum
How to use checksums
0cd052db1fbbf521ed730fe2ac2890934dcc61ef7458f5ec77b628eff61833ed
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via rambo-pypi-grab/0.1.0

Security for zambo-smolagents 0.1.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page