llama-index-tools-zambo

Zambo: zambo.dev | Add Zambo to your AI: zambo.dev/add | AER-1: zambo.dev/aer1

Zambo is the verifiable work layer for AI. This package is the LlamaHub native tool integration for LlamaIndex: it wraps the Zambo MCP endpoint and offline AER-1 receipt verification in the standard BaseToolSpec idiom, so any LlamaIndex agent can call Zambo tools and check verifiable receipts without leaving the framework.

What Zambo gives a LlamaIndex agent

• 153 tools behind one MCP endpoint, https://zambo.dev/mcp, with 35 tools on the default surface.
• HANDOVER, the flagship: wrap up a session in one AI and pick it up in another, with the work, memory, and context carried across.
• A verifiable receipt for every job. A receipt proves the saved result was not changed after the fact. It does not prove the answer was right.
• Offline receipt verification against AER-1, the IETF draft at draft-zambo-aer1-26. The conformance corpus lives at https://zambo.dev/aer1/corpus.json (126 vectors, 129 checks).

Free tier: 20 calls + 1 wrap-up free every day; picking up in another AI is always free. Day Pass: $0.99 for unlimited use for 24 hours. No account and no API key are needed for the free tier.

For one-line AER-1 receipts collected from a running LlamaIndex agent, see the companion package aer1-llamaindex.

Install

pip install llama-index-tools-zambo

Use it

from llama_index.tools.zambo import ZamboToolSpec

spec = ZamboToolSpec()
tools = spec.to_tool_list()

for tool in tools:
    print(tool.metadata.name)

The tool spec exposes three functions:

• list_tools: list the tools the Zambo MCP endpoint offers.
• call_tool(tool_name, arguments_json): call any Zambo tool by name through the MCP endpoint. Every call returns a verifiable receipt for the job.
• verify_receipt(receipt_json): verify an AER-1 workflow receipt offline. The Merkle root is recomputed from the receipt's own steps and the required fields are checked locally, with no network call and no trust in any server.

Example: verify a receipt offline

import json
from llama_index.tools.zambo import ZamboToolSpec, merkle_root

steps = [
    {"seq": 1, "receipt_id": "11111111-1111-4111-8111-111111111111", "tool": "search"},
    {"seq": 2, "receipt_id": "22222222-2222-4222-8222-222222222222", "tool": "summarize"},
]
receipt = {
    "type": "verifiable-workflow-receipt",
    "workflow_id": "33333333-3333-4333-8333-333333333333",
    "receipt_id": "44444444-4444-4444-8444-444444444444",
    "merkle_root": merkle_root([s["receipt_id"] for s in steps]),
    "steps": steps,
}

spec = ZamboToolSpec()
print(spec.verify_receipt(json.dumps(receipt)))
# verified: AER-1 workflow receipt 44444444-4444-4444-8444-444444444444 ...

Links

• Zambo: https://zambo.dev
• Add Zambo to your AI: https://zambo.dev/add
• AER-1 standard: https://zambo.dev/aer1/
• IETF draft: https://datatracker.ietf.org/doc/draft-zambo-aer1/26/
• Verify a receipt: https://zambo.dev/verify/

License: Apache-2.0