This project is a reference implementation for a Trusted Autonomy T&E assurance-case runtime. It is designed to make autonomy/AI evaluation artifacts more traceable, reviewable, and evidence-backed.

It is not a certified safety system, production command-and-control system, weapons-control system, flight-critical runtime, autonomous vehicle controller, cybersecurity authorization package, or substitute for qualified human review.

The official repository is:

IX-Autonomy-Assurance-Case-Runtime

Use repository provenance, commit history, release signatures when available, and maintainer identity checks before trusting any copy, mirror, binary, or ZIP archive.

Do not run unofficial binaries or installer-style downloads claiming to represent this project unless their provenance is independently verified.

The current implementation provides:

• deterministic evidence records,
• SHA-256 evidence bundle hashing,
• hash-chained run ledger entries,
• strict reference validation,
• scenario-catalog validation,
• safety-gate decision records,
• degradation assessment records,
• independent verification summaries,
• machine-readable reports,
• human-readable reports,
• CLI validation paths.

The current implementation does not provide:

• cryptographic signatures,
• public-key infrastructure,
• hardware security module integration,
• immutable storage backend,
• transparency log,
• network service authentication,
• classified-data handling,
• production authorization workflow,
• formal safety certification,
• formal accreditation.

Please report security issues privately to the maintainer instead of opening a public issue with exploit details.

Include:

• affected file or command,
• reproduction steps,
• expected behavior,
• actual behavior,
• whether data integrity, command execution, evidence validation, or ledger integrity is affected.

Do not place classified, controlled, export-restricted, proprietary, personal, credential, key, token, or operationally sensitive data into example evidence bundles or reports.

The examples in this repository are synthetic and intended for software validation only.

Treat evidence and ledger outputs as integrity aids, not proof of absolute truth.

A valid hash chain means the local records are internally consistent. It does not prove that the underlying telemetry, operator input, source system, test environment, or reviewer action was truthful.

Do not use this project to authorize real-world autonomous action without qualified engineering review, safety review, cybersecurity review, legal review, operational approval, and domain-specific test authority.