An engine for modelling and managing decisions.
Find a file
tegwick c93628a017 Let supersede() reissue an expired approval given a fresh window
expired stays a terminal exit from valid (state machine unchanged),
but supersede() now also accepts it as a source status when the
caller supplies a validity window, linking a new successor approval
to the expired original instead of requiring a caller to mint an
unlinked replacement from scratch. Without a fresh window, an expired
source still conflicts exactly as before.

The original's status becomes superseded (not left as a bare dead-end
expired) and superseded_by points at the successor, so the link is
documented rather than hidden -- informed-decision's overview already
renders this automatically, no changes needed there.

Prompted by the operator missing infd-20260927-b01's approval window
during today's key-cape login outage (APPROVAL-WP-0003).

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Assistant: claude-code
Assistant-Model: sonnet
Assistant-Process: 169987@bnt-lap001
Assistant-Session: 322ef1ef-9048-4021-8570-b6d6f6347999
2026-09-27 20:23:56 +02:00
approval_engine Let supersede() reissue an expired approval given a fresh window 2026-09-27 20:23:56 +02:00
deploy Record T03 review admission and measured CPU request 2026-09-14 02:47:31 +02:00
docs Record T03 review admission and measured CPU request 2026-09-14 02:47:31 +02:00
examples Enforce declared human controls at approval binding 2026-09-10 19:26:12 +02:00
history Align to security layer model v0.7 and open the engine spine 2026-08-29 11:58:28 +02:00
intakes Record GH-DEC-2026-005; strike the spent G3 revisit trigger 2026-09-06 01:36:03 +02:00
schemas Enforce declared human controls at approval binding 2026-09-10 19:26:12 +02:00
tests Let supersede() reissue an expired approval given a fresh window 2026-09-27 20:23:56 +02:00
workplans Let supersede() reissue an expired approval given a fresh window 2026-09-27 20:23:56 +02:00
.custodian-brief.md chore(consistency): sync task status from DB [auto] 2026-09-14 01:10:28 +02:00
.gitignore Implement the engine spine: claim, outbox, machine, API 2026-08-29 12:52:49 +02:00
.repo-classification.yaml Add repo classification to close C-24/C-35 2026-08-29 14:43:52 +02:00
AGENTS.md Finish approval engine spine 2026-09-01 23:45:48 +02:00
cadence.yaml Finish approval engine spine 2026-09-01 23:45:48 +02:00
Containerfile Adopt Alpine as the sanctioned base; release candidate scans clean 2026-09-06 23:01:41 +02:00
INTENT.md Apply GH-DEC-2026-017: INTENT.md governs, layer.yaml is derived, no version 2026-09-21 07:35:00 +02:00
layer.yaml Apply GH-DEC-2026-020: remove companion_version from layer.yaml 2026-09-21 09:37:12 +02:00
Makefile Deploy approval service and verify native persistence and audit 2026-09-14 01:08:50 +02:00
pyproject.toml State pdp_digest explicitly; decline to publish a vocabulary mapping 2026-09-06 09:32:11 +02:00
README.md Harden the PEP harness and KeyCape registration request 2026-09-02 15:46:06 +02:00
SCOPE.md Let supersede() reissue an expired approval given a fresh window 2026-09-27 20:23:56 +02:00
WORK-RECORDS.md Sync approval deployment completion index 2026-09-14 01:11:44 +02:00

approval-engine

The approval as a durable, authenticated, consumable object — issued before an action, verified at the moment of use, and provably not replayable.

An Engine, role PIP, in the NetKingdom security layer model (statute v0.7, accepted; operative form net-kingdom/SECURITY-COMPANION.md). It answers one question, totally and decidably:

Is this approval valid right now — for this exact action, target, actor, and purpose — and has it already been used?

It does not decide whether the action is permitted. That is access-engine, which stays NetKingdom's only policy decision point. An approval is one input to that decision.

Deliberately small, boring, and strict: atomic supersession and single consumption are what make Canon test T-06 — Approval Replay passable. Flexibility here would be a defect. Graded, evidence-based progression belongs to maturity-engine; the two engines are deliberate opposites.

See INTENT.md and SCOPE.md. Declaration: layer.yaml. Claim: docs/approval-claim.md. Consume: docs/approval-consumption.md. Caller auth: docs/caller-authentication.md. PEP sequence: docs/pep-integration.md. Origin: flex-auth FLEX-DEC-2026-001, raised while assenting to the security layer model.

make test
approval-engine migrate --db approvals.sqlite
approval-engine verify --db approvals.sqlite
approval-engine serve --db approvals.sqlite \
  --jwt-issuer https://auth.netkingdom.local \
  --jwt-audience approval-engine \
  --jwks-url http://key-cape.sso.svc.cluster.local:8080/jwks

POST /v1/approvals/{id}/consume implements GH-DEC-2026-003: the PEP atomically spends the approval before the protected side effect. Same-digest retries are idempotent; a different digest conflicts.

Production operations, caller scopes, audit delivery, and PEP sequencing are documented in docs/storage-operations.md, docs/caller-authentication.md, docs/outbox-contract.md, and docs/pep-integration.md. The checked-in StatefulSet deliberately refuses production startup without a migrated persistent store, KeyCape JWT verification, and authenticated audit delivery.