Agent Scope Credential (ASC)
User Journey
The Agent Scope Credential defines and verifies the operational boundaries and authorized action space within which an AI agent is permitted to operate. This credential establishes clear, cryptographically-enforced limits on agent behavior, including transaction volumes, operational domains, temporal constraints, and resource access permissions, creating a verifiable "sandbox" for autonomous action.
See It in Action
COMING SOON
Why Verify Agent Scope
The Unbounded Risk Problem
Autonomous agents without defined boundaries create systemic risks:
Mission Creep: Agents expanding beyond intended functionality
Resource Exhaustion: Uncontrolled consumption of computational or financial resources
Cascading Failures: Single agent actions triggering systemic impacts
Regulatory Violations: Operating in unauthorized domains or jurisdictions
The Scope Enforcement Imperative
Verified scope credentials provide:
Predictable Behavior: Clear understanding of agent capabilities and limitations
Risk Containment: Hard limits on potential negative impact
Regulatory Compliance: Demonstrable constraints for compliance purposes
Trust Calibration: Appropriate level of trust based on verified boundaries
Why zkMe ASC
Privacy-Preserving Scope Verification
Zero-Knowledge Proofs: Verify scope compliance without exposing sensitive operational details
Selective Disclosure: Prove specific scope attributes without revealing full constraints
Competitive Protection: Maintain confidentiality of business logic while demonstrating compliance
Technical Superiority
Cryptographic Enforcement: Tamper-proof scope definitions and verification
Real-Time Validation: Sub-50ms scope verification for high-frequency operations
Dynamic Updates: Secure modification of scope parameters with proper authorization
Cross-Platform Consistency: Uniform scope enforcement across different environments
Enterprise-Grade Features
Granular Controls: Multi-dimensional scope definitions (financial, temporal, geographic, functional)
Hierarchical Scoping: Nested scope credentials for complex organizational structures
Audit Trail: Immutable record of scope definitions, modifications, and violations
Integration Ready: SDKs and APIs for seamless platform integration
How It Works
For Agent Developers & Principals:
Scope Definition: Specify operational boundaries based on risk assessment and use case requirements
Credential Creation: Generate scope credentials with defined limits and conditions
Agent Configuration: Embed scope enforcement mechanisms within agent architecture
Scope Attestation: Obtain third-party validation of scope constraints and enforcement
Dynamic Adjustment: Update scope credentials as operational needs evolve
For Platforms & Resource Providers:
Scope Verification: Check agent scope credentials before granting access
Policy Enforcement: Configure systems to respect scope limitations
Monitoring & Compliance: Continuously validate agent actions against attested scope
Automated Enforcement: Implement hard limits based on scope credentials
For Users & Regulators:
Scope Transparency: Review agent operational boundaries before engagement
Risk Assessment: Evaluate potential exposure based on defined limits
Compliance Verification: Ensure agents operate within authorized parameters
Incident Analysis: Reference scope credentials during dispute resolution
Scope Definition Framework
Scope Dimensions → Constraint Definition → Credential Issuance → Enforcement Hooks → Verification ProofsKey Scope Dimensions
Financial Constraints
Transaction Volume Limits: Maximum single transaction value
Cumulative Exposure: Total financial exposure across operations
Asset Restrictions: Approved cryptocurrencies, tokens, or financial instruments
Counterparty Limits: Maximum exposure to individual counterparties
Operational Boundaries
Domain Restrictions: Authorized operational domains (DeFi, CEX, gaming, etc.)
Geographic Limits: Permitted jurisdictions and restricted regions
Temporal Constraints: Operational hours, expiration dates, renewal requirements
Resource Limits: Computational, storage, and bandwidth allocations
Functional Permissions
API Access: Approved endpoints and method permissions
Smart Contract Interactions: Authorized contract addresses and function calls
Data Access: Permitted data sources and usage rights
Tool Usage: Approved external tools and services
Risk Parameters
Volatility Exposure: Maximum allowed market exposure
Leverage Limits: Maximum leverage ratios for trading operations
Position Sizing: Maximum position sizes relative to portfolio
Risk Thresholds: Automated deactivation triggers
Technical Implementation
Credential Structure
{
"scopeId": "urn:uuid:scope-7d8e9f0a...",
"agentDID": "did:agentry:0x1234...",
"issuerDID": "did:agentry:principal:abc123",
"scopeVersion": "2.1.0",
"constraints": {
"financial": {
"maxTransaction": 50000,
"maxDailyVolume": 250000,
"allowedAssets": ["ETH", "BTC", "USDC", "DAI"],
"maxCounterpartyExposure": 100000
},
"operational": {
"allowedDomains": ["defi_lending", "dex_trading"],
"restrictedDomains": ["derivatives", "gambling"],
"geographicLimits": ["US", "EU", "UK"],
"operationalHours": "00:00-23:59 UTC",
"expiration": "2025-12-31T23:59:59Z"
},
"functional": {
"approvedAPIs": ["uniswap/v3", "aave/v3", "compound/v3"],
"allowedContracts": ["0xabc123...", "0xdef456..."],
"maxGasPerTransaction": 5000000,
"dataAccessPermissions": ["public", "user_provided"]
},
"risk": {
"maxVolatilityExposure": 0.15,
"maxLeverage": 3.0,
"maxPositionSize": 0.1,
"riskDeactivationTriggers": ["liquidation_risk > 0.8"]
}
},
"enforcementMechanisms": {
"onChainVerification": true,
"offChainMonitoring": true,
"automatedShutdown": true
},
"proofs": {
"principalAuthorization": "zkp_principal_123...",
"scopeIntegrity": "zkp_scope_456..."
}
}Enforcement Architecture
Pre-Action Verification
Agent generates zero-knowledge proof of scope compliance before execution
Platforms verify proof against current scope credentials
Actions rejected if outside authorized scope
Runtime Monitoring
Continuous validation of agent actions against scope constraints
Real-time alerts for scope boundary approaches
Automated intervention for scope violations
Post-Action Auditing
Immutable logging of all actions with scope compliance status
Regular compliance reporting and anomaly detection
Scope violation analysis and response protocols
Verification Flow
Scope Proof Request: Platform requests proof of specific scope compliance
Zero-Knowledge Generation: Agent generates proof without revealing full constraints
Cryptographic Validation: Proof verified against on-chain scope credentials
Action Authorization: Platform grants or denies access based on scope compliance
Key Benefits
For Agent Principals & Developers
Risk Management: Clearly defined operational boundaries reduce liability exposure
Compliance Demonstration: Verifiable evidence of operational constraints for regulators
Investor Confidence: Transparent risk controls attract funding and partnerships
Operational Efficiency: Automated scope enforcement reduces manual oversight
For Platforms & Service Providers
Systemic Risk Reduction: Prevent agent actions from causing platform-wide issues
Regulatory Compliance: Demonstrate proper oversight of third-party agents
Resource Protection: Prevent resource exhaustion through enforced limits
User Protection: Ensure agents operate within expected parameters
For End Users
Transparent Operations: Clear understanding of agent capabilities and limitations
Risk Awareness: Informed consent about potential exposures and constraints
Recourse Mechanisms: Clear accountability for scope violations
Trust Calibration: Appropriate level of trust based on verified boundaries
For Regulators & Auditors
Oversight Efficiency: Standardized framework for monitoring agent operations
Compliance Verification: Automated validation of operational constraints
Incident Investigation: Clear scope definitions for violation analysis
Market Stability: Reduced systemic risk through contained agent operations
Use Cases to Benefit
DeFi & Financial Services
Trading Agents: Volume limits, asset restrictions, and risk parameter enforcement
Lending Protocols: Borrowing limits, collateral requirements, and liquidation triggers
Yield Farming: Strategy constraints, impermanent loss limits, and protocol approvals
Portfolio Management: Allocation limits, rebalancing thresholds, and risk controls
Enterprise Automation
Procurement Agents: Spending limits, vendor approvals, and contract value constraints
HR Systems: Salary band enforcement, benefit allocation limits, and compliance boundaries
IT Operations: Resource allocation limits, change management approvals, and access controls
Customer Service: Response escalation thresholds, refund authorization limits, and data access permissions
Healthcare Applications
Diagnostic Support: Confidence threshold enforcement and specialist referral triggers
Treatment Planning: Protocol adherence, medication dosage limits, and intervention approvals
Patient Monitoring: Alert escalation criteria and emergency response triggers
Research Operations: Data usage limitations and experimental protocol boundaries
Supply Chain & Logistics
Inventory Management: Reorder thresholds, supplier allocation limits, and quality controls
Logistics Optimization: Route constraints, carrier selection criteria, and cost limits
Procurement Agents: Budget adherence, supplier diversity requirements, and sustainability criteria
Demand Forecasting: Prediction confidence bounds and adjustment limitations
Government & Public Sector
Resource Allocation: Budget enforcement, eligibility criteria, and distribution limits
Regulatory Compliance: Inspection frequency, enforcement actions, and penalty calculations
Public Services: Benefit calculation boundaries, service eligibility, and escalation protocols
Infrastructure Management: Maintenance scheduling, resource allocation, and emergency response
Consumer Applications
Personal Finance: Spending limits, investment allocations, and risk tolerance enforcement
Smart Home: Device control permissions, energy usage limits, and privacy boundaries
Content Creation: Copyright compliance, content guidelines, and distribution limits
Social Media: Posting frequency, content moderation, and interaction limitations
Cross-Domain Applications
Multi-Agent Systems: Inter-agent communication limits and collective action constraints
Federated Learning: Data access boundaries and model update limitations
Autonomous Systems: Operational domain restrictions and safety parameter enforcement
Research Agents: Experimental scope, data usage limits, and publication constraints
Pricing & Integration
Drop us a line at [email protected] and let’s kick things off!
Last updated