deerflow

DeerFlow + xProof

Add on-chain certification to DeerFlow agents via XProofDeerFlowSkill.

What gets certified

Any content the agent passes to the skill — plain text or structured JSON — is certified with:

• WHO — your agent name
• WHAT — SHA-256 hash of the content
• WHEN — UTC timestamp
• WHY — your configured reason

Install

pip install xproof

Usage

from xproof import XProofClient
from xproof.integrations.deerflow import XProofDeerFlowSkill

client = XProofClient(api_key="pm_...")   # get key at xproof.app
skill = XProofDeerFlowSkill(client=client, agent_name="research-agent")

# The agent calls the skill to certify its output
result = skill._run("Q3 revenue grew 15% YoY driven by AI product adoption.")
print(result)  # JSON with proof_id, file_hash, transaction_hash, status

Structured JSON input with custom metadata:

import json

result = skill._run(json.dumps({
    "content": "Market analysis: AI sector growing 40% annually",
    "file_name": "market-analysis.json",
    "author": "market-analyst",
    "why": "Annual market review certification",
}))

Run the demo

python main.py

No API key required — the demo uses a mock xProof client.

Links

• xproof.app
• PyPI: xproof