Quantum Computing Governance Framework: Treating Physical Hardware Fidelity as a First-Class Constraint Dimension in Agent Workflows

aeoess-quantum-governance is an innovative quantum computing governance system. It enforces physical quality constraints before execution by querying IBM Quantum hardware calibration data in real time, introduces hardware fidelity as a core governance dimension for autonomous agent workflows, and generates cryptographically signed execution credentials. This system aims to address the problem where quantum hardware noise limits large-scale applications, ensuring the reliability of the agent execution environment.

Traditional agent governance focuses on dimensions such as budget control and permission scope, but it falls short when dealing with quantum circuit execution. Physical properties of quantum hardware, such as coherence time and gate operation fidelity, change dynamically over time, and performance varies significantly across different qubits and time points. Core insight: In quantum scenarios, the physical quality of hardware must be a core input for governance decisions; otherwise, even if the logic is correct, the execution results from noisy hardware may be meaningless.

The system builds a four-layer quality control architecture:

1. Budget Dimension Check: Verify whether the number of samples, circuit depth, number of qubits, and target backend are within the authorized scope.
2. Real-Time Calibration Data Acquisition: Query real-time data from the IBM Quantum platform, such as T1 relaxation time, T2 decoherence time, gate operation error rate, and readout error rate.
3. Physical Fidelity Check: Compare the physical requirements of the delegation authorization with real-time calibration data to ensure all constraints are met (e.g., T1/T2 thresholds, error rate limits, calibration freshness).
4. Signed Credential Generation: After execution, use the Ed25519 algorithm to generate a cryptographically signed credential containing governance decisions, calibration snapshots, execution results, delegation chains, and timestamps.

The system adopts a monotonically narrowing delegation model. QuantumDelegation includes two types of constraints:

• Budget Dimension: Maximum number of samples, circuit depth, number of qubits, allowed backend list.
• Physical Dimension: Minimum T1/T2 time, maximum readout/gate operation error rate, valid duration of calibration data. Sub-delegations can only be stricter than parent delegations to ensure secure permission transfer. Planning agents can set stricter physical requirements, and execution agents cannot relax them.

Technical Implementation: Developed in Python with a modular structure including delegation.py (delegation constraint check), calibration.py (IBM Quantum data interaction), gateway.py (four-layer process orchestration), receipt.py (Ed25519 signature), and demo.py (demonstration scenarios). Demonstration Scenarios:

1. Bell State Execution (PERMITTED): Budget and fidelity checks passed.
2. Excessive Sampling Request (DENIED_BUDGET): Rejected due to exceeding the authorized number of samples.
3. Strict Physical Requirements (DENIED_FIDELITY): Rejected because the hardware T1 does not meet the minimum threshold.
4. Hybrid Workflow (PERMITTED): Verify the correctness of the monotonically narrowing delegation chain.

Academic Support: The paper Physics-Enforced Delegation: Governing Quantum Hardware Quality in Autonomous Agent Workflows is published on Zenodo (DOI: 10.5281/zenodo.19478584). The project uses the Apache-2.0 license and belongs to the AEOSS ecosystem. Application Value: Ensure the quality of computing results for developers and avoid resource waste from noisy hardware; provide a standardized governance mechanism for platform operators; extend to physical constraint scenarios such as edge computing and IoT.

aeoess-quantum-governance is an important advancement in the intersection of agent governance and quantum computing, emphasizing the importance of the physical reality of underlying hardware. By incorporating real-time calibration data into governance decisions, it provides a foundation for quality assurance of quantum applications. In the future, it will expand to scenarios such as resource-constrained edge computing and IoT energy management, and the physics-aware governance model will become a key infrastructure for reliable quantum applications.