How Treeship Works
Transform any AI agent into a verifiable, trustworthy system with mathematical guarantees
Explore ZK Featuresfrom treeship import zk
# Add ZK attestations to any function
@zk.attest
def process_content(data):
# Your agent logic here
result = ai_model.analyze(data)
# Automatic proof generation:
# ✓ Skill verification
# ✓ Performance metrics
# ✓ Privacy preservation
return result
# Deploy with verifiable capabilities
zk.deploy_with_proofs([
"data-analysis",
"content-processing",
"sentiment-analysis"
])Instant ZK Integration
Set Up in Minutes
Add @zk.attest decorator to any function for instant attestation
Automatic Proofs
Generate verifiable proofs for execution, performance, and capabilities
Skill Certification
Attest to specific AI capabilities with cryptographic guarantees
Privacy by Design
Process sensitive data while maintaining complete privacy
How Proof of Skills Works
Our automated system verifies agent capabilities through documentation analysis, human validation, and real-world performance monitoring
Documentation Scan
We analyze your agent's documentation, code comments, and declared capabilities to understand intended functionality
Treeship Validation
Our agentic validation system automatically tests your agent's outputs, validates claimed capabilities, and tracks performance patterns
Live Performance Tracking
Once deployed, we continuously monitor task completion, multi-agent workflows, and success rates to maintain attestations
Zero-Knowledge Attestations
All performance data and skill validations are processed through zero-knowledge proofs, ensuring your agent's sensitive data and proprietary logic remain completely private while still providing verifiable attestations to third parties.
For Developers
Create Custom ZK-Powered Skills & Attestations
Build unique AI capabilities with built-in verification. Our SDK lets you create custom skills that prove their performance while maintaining complete privacy.
from treeship import ZK, Agent
# Create verifiable AI agent
@ZK.attest_skills(["analysis", "reasoning"])
class AnalysisAgent(Agent):
def analyze(self, data):
# ZK proofs generated automatically
result = self.llm.process(data)
return self.verify_output(result)
# Deploy with attestations
agent = AnalysisAgent()
agent.deploy_verified({
"privacy_level": "high",
"proof_generation": "automatic",
"skills_certified": True
})
Independently Verifiable Credentials
Third parties can verify AI agent capabilities without trusting the developer
Learn MoreZero-Trust Verification
Verify claims without relying on the agent creator's word
Cryptographic Proofs
Mathematical guarantees that cannot be faked or manipulated
Real-Time Validation
Instantly verify current performance and capability claims
Audit Trail
Complete history of agent behavior and performance over time
All claims independently verified via zero-knowledge proofs