Build with AIEP
All core AIEP components are open source under Apache 2.0. Each repository has a specific role in the AIEP system. This page maps them by function so you can find the right starting point.
Try the Verification Playground → — see the protocol in action before you build.
Verification
Validates AI outputs against AIEP schemas, evidence, and hashes. Use this to check whether a response conforms to the AIEP protocol — including evidence rail structure, content hashes, and schema validity.
- Validates
response.json,evidence.json, andevidence-rail.jsonartefacts - Reports pass/fail with detail on which invariants failed
- Used by Piea and Forecast as the canonical verification engine
Source Creation
Creates machine-readable, verifiable content using AIEP Mirror endpoints. Publish a /.well-known/aiep/ layer on any domain in minutes — five JSON files, no backend, no registration required.
- Makes your domain discoverable and citable by AI retrieval systems
- Every published artefact is hash-bound and schema-conformant
- Supports federation with other AIEP Mirror nodes
Integration
The core integration SDK. Implements the AIEP genome specification and enables AIEP integration into applications. Available as aiep-genome-sdk on PyPI.
- Frozen kernel R1–R8 canonical primitives
- Evidence ledger, reasoning ledger, and divergence detection
- Constitutional enforcement — fail-closed execution control
- Dual-ledger memory substrate
Testing
Canonical test vectors for deterministic validation and reproducibility. Use these to verify that your AIEP implementation produces byte-identical outputs for known inputs.
- Canonical evidence artefacts for schema conformance testing
- Hash-verified reference outputs
- Interoperable across all AIEP SDK versions
Shared Tooling
Utility functions and helpers for AIEP ecosystem development. Includes hashing, schema validation, and artefact inspection tools.
Hub
The public hub — this site. Protocol pages, machine-readable endpoints, examples, schemas, and discovery surfaces. Use it as the reference implementation for a well-known endpoint structure.
How they connect
Your application
↓
GENOME-SDK ← integration layer
↓
VALIDATOR ← verify outputs
↑
MIRROR ← publish verifiable sources
↑
VECTORS ← test determinism
↑
UTILITIES ← shared tooling
Machine-readable endpoints
| Resource | URL |
|---|---|
| Protocol discovery | /.well-known/aiep/ |
| Verified response example | /examples/response.json |
| Evidence rail example | /examples/evidence-rail.json |
| Audit trail example | /examples/audit-trail.json |
| Response schema | /schema/response.json |
| Evidence schema | /schema/evidence.json |
| Evidence rail schema | /schema/evidence-rail.json |
Try it first
Verification Playground → — run a live verification end-to-end
See also: What is AIEP? · How AIEP Works · Mirror · Genome SDK · Downloads