Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YWBMovyFoy9RRrfL7zKvPJ
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 4014535@bnt-lap001
Assistant-Session: d0036016-73e8-4da1-8e47-563e3ab39a3c
Regenerated by fix-consistency; adds the inbound v0.3 review intake.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 2564823@bnt-lap001
Assistant-Session: 2a7ed827-4928-4b9f-8613-9135c9cadfe9
Regenerated by fix-consistency; adds the inbound assent intake.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 2564823@bnt-lap001
Assistant-Session: 2a7ed827-4928-4b9f-8613-9135c9cadfe9
risk-nexus accepted the offer to match their convention rather than grow a second
one, and published it: 14d critical/high, 30d medium, 60d low, nothing auto-closing
on staleness alone. Their preference — point warden route gaps at those windows and
the two registers agree without a shared mechanism — is better than a joint tool.
blocker_stale_days() now maps lane risk onto those windows. A flat 14 would have
been wrong in both directions: too aggressive for a low-risk pointer, and it treated
an admin PAT lane the same as one.
ungraded takes the shortest window, not the longest. ADR-0007 makes an absent grade
a defect and ADR-0008 makes a grade cover the whole path, so a lane nobody graded is
the one whose blocker deserves least trust. Encoding that as 60 days would have been
the fail-open default this repo already fixed once.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
secrets-engine reviewed our catalog metadata while drafting their five entries
and graded issue-core-ingestion-api-key and reuse-surface-hub-write-token high.
We had both as standard, and had deliberately regraded them DOWN on 2026-08-19.
They are right. Both paths carry a second credential our grade never looked at --
GITEA_BACKEND_TOKEN (CCR-2026-0002, a deliberate field-set decision) and a
dual-consumer webhook HMAC (CCR-2026-0005). Neither is recovered by rotating the
credential the lane is named after.
The defect is structural: we graded the lane by its headline field, but a read
returns every field at the path. Worse, the evidence was already in the CCRs we
cite as authoritative -- not missing, unread -- and a test asserted the wrong
answer, so a correct first-pass grade got overruled by it.
ADR-0008 records the rule: a grade covers every field its path discloses.
ADR-0007 is unchanged and still governs; this says what the grade is of.
Six of the remaining standard lanes have no KV path. Two have paths and no field
evidence; per ADR-0008 they are stated as unknown rather than assumed, and left
for operator-sanctioned grading.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
agent-high-risk-boundary denies 5 data paths, covering 6 of ops-warden's 17
high-risk lanes. Eight high-risk lanes with concrete KV paths are not denied,
four of which were already graded high before the 2026-08-19 regrade — the
divergence is pre-existing and the regrade only made the lists comparable.
This is the layer that matters most: warden access exits 7 for all 17, but that
protects only the ops-warden path. The policy protects a direct bao kv get,
which is the 2026-07-16 vector.
Routed to risk-nexus as RISK-F-0004 with fix_owner railiance-platform, since the
policy is theirs. Live confirmation outstanding — ops-warden's OpenBao token is
expired, so this is a static file-vs-catalog comparison and the finding says so
rather than implying it was verified against the server.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The mechanism behind RISK-F-0003 was sharper than the finding described.
is_high_risk was risk == "high", but risk was never absent at the model layer:
RouteEntry.risk carried a dataclass default of "standard". An omitted grade was
not unhandled, it was actively resolved to the permissive value — fail-open by
construction, which is why nothing ever warned.
The default is now "ungraded" and is_high_risk returns true for anything outside
an explicit low-risk vocabulary (standard / low / accepted). An omitted grade and
an unrecognised grade from a newer catalog both resolve to high, so the boundary
fails safe in both directions rather than reading an unknown value as permission.
test_every_repo_catalog_lane_is_explicitly_graded is the CI gate that stops an
ungraded lane being committed, per ADR-0007: absence is not a grade.
"accepted" is in the low-risk vocabulary deliberately, ready for the
maturity-derived default — an experimental-context lane may be explicitly
accepted, which is a graded decision rather than an omission.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Closes the RISK-F-0003 exposure. All 14 ungraded lanes now carry an explicit
risk grade with its justification in the entry: 17 high, 10 standard, 0
ungraded. The agent read-boundary now fires (exit 7) on lanes that were silently
outside it.
Graded on merit rather than defensively. A first pass marked two ordinary
internal workload secrets high; test_high_risk_lanes_classified asserted the
opposite and was right, so both were regraded down. high means disclosure into a
logged context is damaging beyond what rotation recovers. inter-hub-bootstrap-ssh
is high conservatively, with the reason in the entry so it is regraded with
evidence rather than assumed down.
ADR-0007 records the rule the grading rests on: build-stage permissiveness
applies to controls that gate work, not to controls that prevent credential
disclosure. The test is friction, not severity — the read-boundary blocks nobody,
since --out/--exec/--wrap remain available, so relaxing it buys nothing. A
blocked operation is recovered by retrying; a credential in a logged transcript
is not recovered by rotation.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
zone-engine is seeded and owns the security zone model as ZONE-WP-0001. Under
ADR-0005 ops-warden implements one lane narrowly and routes the rest, and an
estate-wide enforcement model is not a lane to absorb — it was ops-warden's
deferred flip that exposed the gap, not ops-warden's model to define.
WARDEN-WP-0032 is rewritten as the consumer side: hand the estate inputs to
ZONE-WP-0001-T02 (27 catalog lanes, the actor inventory, the three posture axes,
the three controls the model must express, and the compiled-registry path),
then replace policy.enabled with a zone-aware control and amend ADR-0006 to say
ops-warden follows the model rather than owning it.
ADR-0006 and SCOPE updated to point at zone-engine, which joins the related
repositories table.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Expands the placeholder into a real plan, grounded in what the estate already
has rather than first principles:
- tenancy-posture_v0.1 is the structural model to follow (canon standard +
per-repo declaration + "accuracy, not altitude"), and its §14 adoption stance
is "structure, not tooling" — a precedent about sequencing.
- Reefs are substrate placement zones with real risk statements. A security
zone is not a reef; repo-manager's own "topology is not readiness" applies
here as "placement is not posture".
- flex-auth is latency-critical and already consumes a compiled registry
snapshot, so zone membership should reach the PDP by compilation rather than
a synchronous lookup in the decision path.
- warden plan already returns verdicts with reasons; zone-aware enforcement
extends that machinery instead of growing a parallel one.
T01 settles ownership before any modelling, because it decides which repo the
rest lives in: canon standard, a zone-engine runtime authority, or flex-auth
policy data. The recommendation is canon-first with zone-engine named as an
explicit promotion — but the time-boxed refactor exception is the case that
could justify an engine immediately, since an expiring grant is state and state
wants an owner.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
CoulombCore is being retired, so the docs stop using it as the reference host:
state-hub-coulombcore examples become state-hub-railiance01, and the reuse-surface
playbook no longer attributes bao.coulomb.social to it — that resolves to
92.205.62.239, which is railiance01. The openrouter lane keeps its factual note
about ESO on the CoulombCore cluster, with a retirement flag for its owner.
Also corrects this session's own error. The WP-0031 evidence blamed the
Unauthorized on a local-port collision between k3s-api-coulombcore and
k3s-api-haskelseed. That was wrong: the haskelseed tunnel is a reverse forward,
where local_port is a destination rather than a listener, so they never competed.
16443 was simply CoulombCore's k3s — a different cluster whose client CA does not
know that cert. The wrong reason had already gone to flex-auth, so it is
corrected in the file rather than quietly dropped.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
flex-auth's binding names system:serviceaccount:ops-warden:ops-warden, and that
ServiceAccount did not exist. deploy/kubernetes/caller-identity.yaml creates it
plus its namespace — no RBAC, automount off; it is never used to call the
Kubernetes API, only to be TokenReviewed. Applied to the railiance01 cluster.
Operator warden.yaml now uses caller_auth mode: command (kubectl create token,
audience flex-auth, 10m). Gate exits 0 live against a port-forward of the pin:
HTTP 200, effect=allow, decision:f3f7c88f9585582a.
The evidence is not that allow — warn allows anonymous callers too. It is that
the pin's "caller authentication warning" count held at 4 across two
authenticated runs. That is the ADHOC-2026-08-17-T01 condition.
Also gives the readiness probe a structurally complete context, so a deny means
the policy said no rather than the probe being malformed.
policy.enabled stays false. T05 waits on flex-auth setting callerAuth.mode:
enforce (their FLEX-WP-0016 T03).
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Founder decision 2026-08-11: neither withdraw nor proceed. The broker covers
the workstation, but unattended signing on a remote tunnel host is unresolved,
and there the AppRole is narrower than placing the broker's issuer token.
Recorded where each trigger will be hit: WP-0027 break-glass task (trust-root
design), and the ops-bridge cutover playbook (token source must be resolved
before going live). secrets-engine told to stop holding apply readiness.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Founder directive: ops-warden works with, never replaces or duplicates,
secrets-engine / tenant-engine / user-engine. Covering an unfilled gap is
acceptable only as a tracked interim with a named intended owner.
- INTENT §9 "Cover gaps, but never silently own them"; success criterion 7;
tenant-engine and user-engine added to the literacy table; non-goal on
permanently owning another component's lane
- WP-0030 (proposed): delegation: metadata, backfill, warden route gaps,
promotion gate, publish the register to owner repos
- history/2026-08-11-delegation-surface-assessment.md: 2 of 24 lanes carry
exec_owner; 11 proxies record no intended owner
- SCOPE refreshed to 2026-08-11 (was 6 workplans behind); completeness C5 -> C4
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
