Loading AGEI...
Canonicalized → Hashed → Signed → Governed → Watermarked → Released → Verifiable
AGEI ships with a default catalog of 10 base governance policies and gates that prove the core AGEI claim:
Every governed AI event is policy-evaluated, hashed, signed, linked, and verifiable.
Rules, requirements, and default failure actions
Policy decisions at critical lifecycle checkpoints
Signed, verifiable evidence of governance
┌──────────────┐
│ AI Event │ (text prompt, image generation, agent response)
└──────┬───────┘
│
v
┌──────────────────────────────────┐
│ Gate: AI Interaction │ → Outcome: INSPECT
│ Provenance │ Reason: privacy_preserving_model_provenance
└──────┬───────────────────────────┘
│
v
┌──────────────────────────────────┐
│ Gate: Watermark Requirement │ → Outcome: APPROVE
└──────┬───────────────────────────┘ Reason: watermark_required_and_applied
│
v
┌──────────────────────────────────┐
│ Gate: Dual-State Provenance │ → Outcome: APPROVE
└──────┬───────────────────────────┘ Reason: dual_state_hashes_verified
│
v
┌──────────────────────────────────┐
│ Gate: Artifact Release │ → Outcome: APPROVE
│ Authorization │ Reason: release_authorized
└──────┬───────────────────────────┘
│
v
┌──────────────────────────────────┐
│ Gate: Evidence Integrity │ → Outcome: APPROVE
└──────┬───────────────────────────┘ Reason: signed_ed25519_evidence
│
v
┌──────────────────────────────────┐
│ Gate: Signature Verification │ → Outcome: APPROVE
└──────┬───────────────────────────┘ Reason: valid_active_signing_key
│
v
┌──────────────────────────────────┐
│ Gate: Receipt Chain Integrity │ → Outcome: APPROVE
└──────┬───────────────────────────┘ Reason: chain_verification_passed
│
v
┌──────────────────────────────────┐
│ Gate: Vault Commitment │ → Outcome: APPROVE
└──────┬───────────────────────────┘ Reason: vault_committed
│
v
┌──────────────────────────────────┐
│ Verifiable Receipt Published │ ✓ Evidence-grade
│ Public Verification URL │ ✓ Audit-pack eligible
└──────────────────────────────────┘Every gate evaluation produces one of four outcomes
All policy rules passed. Allows the action to proceed.
Policy violation detected. Blocks the action.
Flags for review but allows action. Creates audit trail.
Requires human-in-the-loop (HITL) decision before proceeding.
Click each gate to see its policy rules, outcomes, and reason codes
This is an actual artifact verification from AGEI showing all base gates in action
watermark_required_and_applied
You can create custom policies and gates for your specific use cases, but do not delete or disable the base gates unless you understand the implications:
Best Practice: Leave base gates active and layer your custom policies on top of them.
Policy and gate slugs for reference
The base policy catalog is seeded automatically when you run database migrations. If you need to manually seed:
psql -f supabase/seeds/agei_base_policies_gates.sqlThis creates 1 policy set, 10 policy versions, and 10 gate definitions for your organization.
AGEI base gates are default governance controls that create verifiable evidence of AI governance. They are not legal guarantees, warranties, or compliance certifications. Organizations using AGEI are responsible for configuring policies to meet their regulatory, legal, and ethical requirements.