flex-auth/docs/decision-record-contract.md
tegwick afd9be5aa9
All checks were successful
CI Smoke / host-smoke (push) Successful in 0s
CI Smoke / container-smoke (push) Successful in 3s
Build and Publish Container Image / build-and-push (push) Successful in 57s
fix: the address we published was a misdirection, and the channel is unauthenticated
secrets-engine probed the Service DNS name handed over in FLEX-WP-0021-T05
and found it resolves, from the workstation, to an unrelated public host.
Reproduced here: search ad.binect.de answers wildcard, so
flex-auth-secrets-engine.flex-auth.svc.cluster.local and
this-service-does-not-exist.flex-auth.svc.cluster.local both resolve to
80.158.43.29, while the trailing-dot FQDN correctly fails. A bare Service
name in a handover is not merely unreachable from there, it is a live
misdirection, and the handover was ours.

Had a deployment pointed at it, the CheckRequest body would have gone to
that host: subject, tenant, lane and resource ids, stage, field names,
purpose, plus the caller's bearer token.

Trailing-dot FQDN and "in-cluster only" now replace the bare name in the
example README, SCOPE.md, and the T05 note.

Their real question was how the response channel is authenticated, and
they declined to answer it locally because choosing a transport control
for our service is not a consumer's call. Right boundary, so the answer
is recorded here as FLEX-DEC-2026-010: it is not authenticated. Pins
serve plain HTTP, the envelope carries no signature, and a responder that
knows the package id and version can return a well-formed allow that
passes every check a consumer performs.

The part worth stating in the contract is that the digests do not help
and look like they do. Every input to request_digest,
policy_package_digest and registry_snapshot_digest is either sent by the
caller or published in this repo, so a forger reproduces all three
exactly. They establish integrity of the binding, never authenticity of
the source — and publishing more digests makes a forged envelope look
more authenticated, not less.

For secrets-engine specifically: fail-closed protects against a PDP that
is absent, not against one that lies. An unreachable PDP denies; a lying
PDP allows.

Third instance of one seam in three decisions. 008: a tenant carried
into the digest and never compared — visible, not enforced. 009: a caller
authenticated and never recorded — enforced, not visible. 010: a record
verifiable and unauthentic — checkable, but not evidence.

One nuance that changes the operator recommendation: kubectl port-forward
does authenticate the responder, transitively — no DNS name, one named
pod, API-server TLS. That is the exact reverse of the caller direction,
where it bypasses the NetworkPolicy. Independent properties pointing
opposite ways, so neither can be summarised as "the network protects it".

FLEX-WP-0024 carries signing; key custody routes through warden/OpenBao
rather than minting a key here.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014aQMM1dPXaPiXVn6DwwtLd

Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 715613@bnt-lap001
Assistant-Session: fabd95c1-4c9e-4080-8849-8707ae025f80
2026-09-06 22:44:45 +02:00

83 lines
4.1 KiB
Markdown

# Decision-record contract
Status: published
Contract: `flex-auth.decision-record.v1`
Schema: [`../schemas/decision_envelope.schema.json`](../schemas/decision_envelope.schema.json)
Date: 2026-09-02
This is flex-auth's output artifact under the NetKingdom Security Layer Model
v0.7 §17. Taxonomy holds only the shared field vocabulary. Consumers may rely
on this schema.
A decision record is a `DecisionEnvelope` returned by `POST /v1/check` and the
CLI `check` / `batch-check` / `list-allowed` commands. Standalone evaluation
and every delegated adapter (Topaz, relationship, rule, Keycloak) emit the
same shape.
## Required fields
| Field | Meaning |
| --- | --- |
| `id` | Deterministic decision identifier |
| `effect` | `allow`, `deny`, `redact`, `audit_only`, or `not_applicable` |
| `subject` / `resource` | Normalized refs the evaluator used |
| `provenance` | Who evaluated, over which policy and facts |
## Contract fields consumers may rely on
| Field | Meaning |
| --- | --- |
| `contract_version` | `flex-auth.decision-record.v1` |
| `binding` | Structured subject, action, resource, context, and `request_digest` |
| `binding.approval_binding_digest` | Present only when the request carried `context.approval`. The digest an approval's `pdp_digest` must equal — see [`canonical-request-digest.md`](canonical-request-digest.md). Not a replay identity |
| `lifetime` | **Required on every allow.** A TTL with `not_before` and `expires_at` |
| `provenance.policy_package` / `policy_version` | Named package pin |
| `provenance.policy_package_digest` | SHA-256 of package metadata plus compiled Rego |
| `provenance.registry_snapshot_digest` | SHA-256 of the canonical registry snapshot |
| `provenance.directory_etag` | Directory consistency token when a delegated directory was joined |
| `provenance.input_claim_digests` | SHA-256 per request-time claim class (`context`, `caring_context`) |
| `provenance.decision_time` | UTC timestamp used to compute `lifetime` |
`reason`, `diagnostics`, and CARING prose are not an authorization contract.
## Allow lifetime
Every allow carries `lifetime.kind = ttl`. The duration comes from the policy
package `allow_ttl` field, or from the engine default of `15m` when the package
omits it. A package that declares `allow_ttl: none` (or `0s`) produces a deny
with reason `allow_lifetime_unstated` instead of a standing grant.
Replay is permitted only while `lifetime.expires_at` is still in the future.
See [`canonical-request-digest.md`](canonical-request-digest.md) and
[`decision-input-freshness.md`](decision-input-freshness.md).
## Versioning
This is contract version 1. Additive optional fields may appear. Removing or
redefining a required field requires a new `contract_version` value and a new
schema id.
## What this record does not prove: who answered
**A consumer must not read digest recomputation as verification of the
responder.** Recomputing `binding.request_digest`,
`provenance.policy_package_digest`, and `provenance.registry_snapshot_digest`
and finding all three correct says nothing about who produced the envelope.
Every input to those digests is either sent by the caller or published: the
request material is what the caller just transmitted, and both the package and
registry digests are computable from files in this repo. A responder that knows
the package id and version can reproduce all three exactly. The digests
establish **integrity of the binding**, never **authenticity of the source**.
`flex-auth.decision-record.v1` carries no signature today, and pins serve plain
HTTP. So the response channel is unauthenticated, stated as a stance rather than
left as an assumption (`FLEX-DEC-2026-010`). For a fail-closed consumer the
distinction that matters is this: **fail-closed protects against a PDP that is
absent, not against one that lies.** An unreachable PDP denies; a lying PDP
allows.
A detached signature over the canonical envelope is the intended fix
(`FLEX-WP-0024`). Until it lands, responder authenticity comes from the channel
alone — and of the available channels only `kubectl port-forward` supplies it,
by targeting one named pod over the API server's TLS with no DNS name resolved.