Public promise
Builders can run governed Intelligence and teach and revise a reasoning system through an obvious supported lifecycle: propose, inspect, approve, version, use, measure, revise, roll back, expire, disable, or retire.
Current state
GI1 passed in ace-core 0.4.0. Core + Intelligence ship in one install with governed LIVE source ingress, the LIVE Intelligence bridge, governed Brief synthesis, an inert JSON Domain Pack compiler, exact replay, packaged external-consumer conformance seams, and the unchanged eleven-tool MCP surface.
GC1 passed against public ace-core 0.4.4. An independent Market Intelligence consumer reproduced proposal, human disposition, exact revision use, restart persistence, retirement failure, and attribution against a clean public-package deployment. Released Core conformance tests cover rejection, revision, rollback, expiry, unavailable dependencies, effectiveness classification, and non-selectable revision/retirement proposals.
Closeout: Core PR #69, public evidence, and Market external-consumer PR #3.
Public scope
- Make the canonical recipe, instrument, framework, tool, perspective, and procedural-knowledge lifecycle obvious through supported public interfaces.
- Create inspectable proposals from authorized tasks, corrections, conversations, and documents.
- Preserve immutable revisions, durable human approval receipts, provenance, explicit authority, bounded discovery, and exact use attribution.
- Support rejection, rollback, supersession, conflict, expiry, disablement, and retirement without erasing history.
Acceptance gate
- An external consumer can teach ACE a reusable capability through supported public interfaces.
- A human can inspect the semantic change and approve or reject it with a durable receipt.
- A fresh invocation materially uses the exact approved revision with complete attribution.
- Unauthorized, rejected, unavailable, expired, or superseded cognition fails closed.
- The complete journey is reproducible from public artifacts without requiring ACE internals.
Result: passed.
Explicit boundary
Core owns governed reasoning, durable state, authority, decisions, receipts, and outcomes. Intelligence owns domain-neutral sensing and orientation machinery. Domain Packs own ontology, detector configuration, personas, and synthesis policy; connectors own bounded source translation. Models may propose, but deterministic lifecycle policy and explicit human authority own activation and retirement.
This release grants no writable model authority and introduces no silent self-modification. E1's trusted in-process boundary is not hostile-code sandboxing, a human penetration test, or certification.
Depends on 0.3.x — Productized State. See the 0.4.4 release, public roadmap, and live Project.
Public promise
Builders can run governed Intelligence and teach and revise a reasoning system through an obvious supported lifecycle: propose, inspect, approve, version, use, measure, revise, roll back, expire, disable, or retire.
Current state
GI1 passed in ace-core 0.4.0. Core + Intelligence ship in one install with governed LIVE source ingress, the LIVE Intelligence bridge, governed Brief synthesis, an inert JSON Domain Pack compiler, exact replay, packaged external-consumer conformance seams, and the unchanged eleven-tool MCP surface.
GC1 passed against public ace-core 0.4.4. An independent Market Intelligence consumer reproduced proposal, human disposition, exact revision use, restart persistence, retirement failure, and attribution against a clean public-package deployment. Released Core conformance tests cover rejection, revision, rollback, expiry, unavailable dependencies, effectiveness classification, and non-selectable revision/retirement proposals.
Closeout: Core PR #69, public evidence, and Market external-consumer PR #3.
Public scope
Acceptance gate
Result: passed.
Explicit boundary
Core owns governed reasoning, durable state, authority, decisions, receipts, and outcomes. Intelligence owns domain-neutral sensing and orientation machinery. Domain Packs own ontology, detector configuration, personas, and synthesis policy; connectors own bounded source translation. Models may propose, but deterministic lifecycle policy and explicit human authority own activation and retirement.
This release grants no writable model authority and introduces no silent self-modification. E1's trusted in-process boundary is not hostile-code sandboxing, a human penetration test, or certification.
Depends on 0.3.x — Productized State. See the 0.4.4 release, public roadmap, and live Project.