Module 2: VCP Architecture

2.1 — The Bundle: VCP's Core Unit

A VCP bundle is a signed package containing:

• Content: The constitutional text itself (principles, rules, guidance)
• Manifest: Metadata — who authored it, when, what version, content hash, expiry, token budget
• Signatures: Cryptographic proof from the issuer and (optionally) a safety attestation from an independent auditor

Bundles are immutable. You don't edit a bundle; you issue a new version and revoke the old one. This creates a tamper-evident history of value changes.

Here is an annotated example of a minimal bundle manifest (JSON):

{
  "vcp_version": "3.1",
  "bundle": {
    "id": "b_healthcare_v3",
    "version": "3.0.0",
    "content_hash": "sha256:a1b2c3d4...",
    "content_encoding": "utf-8",
    "content_format": "markdown"
  },
  "issuer": {
    "id": "org:mercy_hospital",
    "public_key": "ed25519:...",
    "key_id": "key_mercy_2026"
  },
  "timestamps": {
    "iat": "2026-02-28T12:00:00Z",
    "nbf": "2026-02-28T12:00:00Z",
    "exp": "2026-08-28T12:00:00Z",
    "jti": "550e8400-e29b-41d4-a716-446655440000"
  },
  "budget": {
    "token_count": 2048,
    "tokenizer": "cl100k_base",
    "max_context_share": 0.15
  },
  "safety_attestation": {
    "auditor": "auditor:ethics_board",
    "auditor_key_id": "key_ethics_2026",
    "reviewed_at": "2026-02-27T10:00:00Z",
    "attestation_type": "FULL_AUDIT",
    "signature": "ed25519:..."
  },
  "signature": {
    "algorithm": "ed25519",
    "value": "...",
    "signed_fields": ["bundle", "issuer", "timestamps", "budget"]
  }
}

2.2 — CSM-1: The Compact State Message

DEV CSM-1 (Constitutional Semantics Mark 1) is VCP's token format — a compact, machine-readable encoding of constitutional identity. Think of it as a fingerprint for a value configuration.

Format: {persona}{level}[+scopes][:namespace][@version]

Personas define the constitutional archetype:

Level (0–5): Intensity of the persona's influence (0 = disabled, 5 = maximum).

Scopes define which life domains the constitution covers:

Examples:

• N5+F+E — Nanny persona at maximum intensity, covering family and education
• Z3+P:SEC — Sentinel at level 3, privacy scope, security namespace
• G4@1.0.0 — Godparent at level 4, version 1.0.0
• M2+W+G — Muse at level 2, work and general scopes

CSM-1 tokens enable quick matching: "Does this AI system support the constitution I want to use?" without transmitting the full constitutional text.

Try it yourself: The VCP Inspector lets you decode and encode CSM-1 tokens interactively. Paste any token into the Decode tab to see its layer-by-layer breakdown, or use the Encode tab to build tokens with live preview.

2.3 — Verification: Trust Without Faith

When an AI system receives a VCP bundle, it doesn't just trust it. The verification pipeline checks:

1. Schema validity — Does the manifest have all required fields?
2. Size limits — Is the manifest/content within acceptable bounds?
3. Issuer trust — Is this signed by a trusted issuer?
4. Signature validity — Does the cryptographic signature verify?
5. Auditor trust — Is the attestation from a trusted auditor?
6. Attestation validity — Does the auditor's signature verify?
7. Content integrity — Does the content match its hash?
8. Temporal validity — Is the bundle within its valid time window?
9. Replay protection — Has this exact JTI been seen before?
10. Scope matching — Does the bundle match the required model/purpose/environment?
11. Budget compliance — Does the content fit within the model's context budget?
12. Revocation status — Has this bundle been revoked?

Use the VCP Inspector's Capability tab to simulate capability negotiation and see how VCP-Hello/VCP-Ack exchanges work.

Result: one of 18 verification states, grouped by severity:

GOVERNANCE This verification model means organisations can prove which values were active during any AI interaction — critical for compliance, audit, and accountability.

2.4 — The Value Lifecycle

End-to-end flow:

Author constitution → Sign bundle → Publish →
   AI system fetches → Verifies → Injects into context →
      AI responds guided by values →
         Audit trail records what was active

At any point: revocation can invalidate a bundle, attestation can add third-party verification, and context adaptation can adjust how the values are applied based on the situation.

VCP is a trust architecture. Values are authored, signed, verified, applied, and audited with cryptographic integrity at every step.