t81-bundle-creator

name	t81_bundle_creator
description	Create T81 decision bundles with complete provenance and cryptographic verification
metadata	{"openclaw":{"os":["darwin","linux"],"requires":{"bins":["t81"]}}}

T81 Bundle Creator Skill

This skill creates T81 decision bundles that package AI decisions with complete provenance, cryptographic verification, and immutable audit trails.

Usage

When the user needs to create decision bundles from T81 execution results, use this skill to package decisions with complete provenance.

Commands

Create decision bundle:

create-bundle <execution_result> [--output <bundle_file>] [--policy <policy_file>] [--include-provenance]

Bundle from execution:

bundle-from-execution <program_file> [--input <input_data>] [--policy <policy_file>] [--bundle-output <bundle_file>]

Bundle information:

bundle-info <bundle_file> [--detailed] [--verify-integrity] [--output <info_file>]

Sign bundle:

sign-bundle <bundle_file> [--key <key_file>] [--certificate <cert_file>]

Verify bundle:

verify-bundle <bundle_file> [--check-signature] [--check-provenance] [--check-integrity]

Merge bundles:

merge-bundles <bundle1> <bundle2> [--output <merged_bundle>] [--conflict-resolution keep|merge|fail]

Extract from bundle:

extract-bundle <bundle_file> [--component result|provenance|metadata] [--output <output_file>]

Examples

# Create bundle from execution result
create-bundle execution_result.json --output decision.bundle --include-provenance

# Bundle directly from program execution
bundle-from-execution inference.t81 --input data.json --policy security.apl --bundle-output decisions.bundle

# Get detailed bundle information
bundle-info decision.bundle --detailed --verify-integrity --output bundle_info.json

# Sign bundle with cryptographic key
sign-bundle decision.bundle --key private_key.pem --certificate cert.pem

# Verify bundle integrity and signature
verify-bundle decision.bundle --check-signature --check-provenance --check-integrity

# Merge multiple bundles
merge-bundles bundle1.bundle bundle2.bundle --output merged.bundle --conflict-resolution merge

# Extract specific component from bundle
extract-bundle decision.bundle --component provenance --output provenance.json

Bundle Structure

Core Components:

• Decision Result: Primary AI decision or computation result
• Provenance Chain: Complete execution trace and audit trail
• Policy Evidence: All policy decisions and justifications
• Metadata: Timestamps, versions, environment information
• Cryptographic Hashes: Content verification and integrity

Optional Components:

• Input Data: Original inputs (if permitted)
• Execution Logs: Detailed step-by-step execution
• Performance Metrics: Timing and resource usage
• Signatures: Cryptographic signatures for verification

Bundle Format

Bundle Directory Structure:

decision_bundle.v1/
├── bundle.json          # Main bundle manifest
├── result.json          # Decision result
├── provenance.json      # Complete provenance
├── policy_evidence.json # Policy decisions
├── metadata.json        # Bundle metadata
├── hashes.json          # Component hashes
├── signature.sig        # Cryptographic signature (optional)
└── components/          # Additional components
    ├── input_data/
    ├── execution_logs/
    └── performance/

Bundle Manifest (bundle.json):

{
  "bundle_version": "v1",
  "bundle_id": "CanonHash81_BUNDLE...",
  "created_at": "2026-04-04T15:30:00Z",
  "creator": "t81_v1.9.0",
  "components": {
    "result": "result.json",
    "provenance": "provenance.json",
    "policy_evidence": "policy_evidence.json",
    "metadata": "metadata.json",
    "hashes": "hashes.json"
  },
  "integrity_hash": "CanonHash81_INTEGRITY...",
  "policy_compliance": "verified",
  "determinism_guaranteed": true
}

Output Format

Bundle Creation Result:

{
  "status": "success",
  "bundle_file": "decision.bundle",
  "bundle_id": "CanonHash81_BUNDLE...",
  "timestamp": "2026-04-04T15:30:00Z",
  "components": {
    "result": "included",
    "provenance": "included",
    "policy_evidence": "included",
    "metadata": "included"
  },
  "verification": {
    "integrity_hash": "CanonHash81_INTEGRITY...",
    "policy_compliance": "verified",
    "determinism_guaranteed": true,
    "cryptographic_secure": true
  },
  "size": {
    "total_bytes": 1048576,
    "compressed": false,
    "components_count": 5
  }
}

Bundle Verification Result:

{
  "bundle_file": "decision.bundle",
  "verification_status": "passed",
  "timestamp": "2026-04-04T15:30:00Z",
  "checks": {
    "integrity": "passed",
    "signature": "passed",
    "provenance": "passed",
    "policy_compliance": "passed",
    "determinism": "verified"
  },
  "details": {
    "bundle_id": "CanonHash81_BUNDLE...",
    "creator": "t81_v1.9.0",
    "created_at": "2026-04-04T15:30:00Z",
    "components_verified": 5,
    "hashes_valid": true
  }
}

Cryptographic Security

Content Hashing:

• SHA-256 for individual components
• Merkle tree for bundle integrity
• CanonHash81 for T81-specific verification

Digital Signatures:

• RSA/ECDSA signature support
• Certificate-based verification
• Key management integration

Integrity Protection:

• Tamper-evidence detection
• Content verification
• Secure storage recommendations

Policy Integration

Policy Evidence:

• All policy decisions recorded
• Rule-based explanations
• Compliance documentation

Access Control:

• Bundle creation permissions
• Component access restrictions
• Sensitive data handling

Audit Requirements:

• Regulatory compliance support
• Complete audit trails
• Immutable record keeping

Bundle Operations

Creation Options:

• Include/exclude specific components
• Compression settings
• Encryption options
• Signature requirements

Verification Features:

• Multi-level integrity checking
• Signature validation
• Provenance verification
• Policy compliance checking

Merge Capabilities:

• Conflict resolution strategies
• Component combination rules
• Metadata consolidation
• Integrity preservation

Error Handling

• Creation Errors: Component validation and guidance
• Verification Errors: Specific failure explanations
• Signature Errors: Cryptographic problem resolution
• Merge Errors: Conflict resolution assistance

Use Cases

Regulatory Compliance:

• Financial decision documentation
• Healthcare audit trails
• Government record keeping

Supply Chain Integrity:

• AI decision provenance
• Model version tracking
• Result verification

Research Reproducibility:

• Experiment documentation
• Result verification
• Method reproducibility

Integration

This skill wraps T81's bundle creation and management:

t81 bundle create <result> --output <bundle> --policy <policy> --include-provenance