Write Secrets with kubectl replace in the rotation script, since apply copies
the data into the last-applied annotation. Record in operations.md the
live-change rules that apply here and the open finding: four live Secrets
carry that annotation, and removing it waits for the founder's go-ahead.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 352750@bnt-lap001
Assistant-Session: de41ef1c-2113-4dd2-9b92-f318ffa7f98b
railiance-platform replied on all three open threads. Recording what they add
rather than what we already knew.
T02: they reached the same reading of the verifier receipt independently and add
human_client_consume_denied: true for both clients, confirm the verifier ran
twice (activation and post-rollout) at generation 38 with the signing key
unchanged, and state T02 can close against that receipt. They also correct our
framing, rightly: the packet says client-side retrieval is unadmitted, which
stays true, but the attended operator path is not a client-side read and never
required one. This task had been treating those as the same constraint.
T04: the live approval clients mean this repo now holds a committed receipt that
looks like rotation evidence and is not. Both owners independently state the same
two gaps -- no real predecessor rotation, no observed wall-clock expiry. Recorded
in T04 rather than only in T02, because verify-client's predecessor rejection is
now implemented and unproven, which is a different state from missing or done,
and T04 is where that distinction belongs.
Answered their open question on CCR-2026-0020, which has no named presenting
actor. As issuer: the registration carries both approval:create and
approval:approve, so one presenter can create an entry and approve it. That is a
separation-of-duties property of the holder, not a defect in the token -- KeyCape
issues the grants approval-engine asked for. Recorded both shapes KeyCape can
support, and that who holds it is approval-engine's decision and the doctrine
question gate-house's, not ours.
Also narrowed the operations note: the deployed config has not been written since
the activation, so the inspected state is the state that will boot.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016uV8zoCKpA1WRAxsKRYbdH
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1182213@bnt-lap001
Assistant-Session: 966597b9-ae61-46a4-8b9e-1594ab3ec4ad
Both changes on main that postdate the deployed image affect startup or issuance
and both fail closed, so a mistake in either presents as a refusal rather than a
warning. That fact currently lives only in a State Hub message to the deployment
owner; messages age out of attention, and the repo is what the next session
reads, so operations.md is where it belongs.
Records what was actually checked rather than what is assumed: the deployed image
digest and the commit it was built from, that the live config's three browser
clients carry neither newly-rejected field, and that no deployed client declares
a tenant on a browser grant so the refusal is not yet reachable. Also states
plainly that neither change has been exercised against a running issuer, and that
the proof belongs to the attended window rather than to a session running against
production on its own.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016uV8zoCKpA1WRAxsKRYbdH
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 1182213@bnt-lap001
Assistant-Session: 966597b9-ae61-46a4-8b9e-1594ab3ec4ad
G10 waits on custody and platform owners and cannot close from here. What was
doable: verify the handoffs actually went out, replace a remembered live state
with an observed one, and find out whether main is safe to deploy. The last
question found a defect in this repository's own recent work.
Handoffs verified independently rather than trusted: all seven messages are in
the hub with receipt ids. This gap was reopened once for claimed-but-unsent
delivery, so the claim deserved the same scrutiny.
Live state read from the cluster read-only: image main-153258b, only the Qonto
secret materialized so the approval clients remain unprovisioned, four registered
clients, no tenantEngine block. That also corrects an earlier claim of mine --
the deployed config sets userOU explicitly, so the KEY-WP-0023 default fix was
never a production issue.
The precondition: KEY-WP-0019 discovers the expected issuer from
authelia.tokenBaseURL, and the deployed Authelia derives its advertised issuer
from the request Host, advertising the in-cluster address to KeyCape and the
browser-facing one to browsers. Verification fails closed, so a mismatch breaks
every human login and looks like a broken login rather than a misconfiguration.
Which value the token carries needs a real login against production to settle and
was not determined here.
Two mitigations: docs/operations.md documents pinning authelia.issuer and
jwksUrl, with the curl that reveals what the provider advertises for a given
Host; and the authentication failure event now carries a specific reason, so
id_token_issuer_mismatch is distinguishable from a signature failure or an
unreachable key set. The browser still learns nothing.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
Closes gap G08. /healthz returned a constant without probing anything, the server
called ListenAndServe with no signal handling, and the operational limits of
in-memory state, startup-loaded keys and local-only logout lived in code comments
rather than anywhere an operator would look.
/readyz probes LLDAP, Authelia and privacyIDEA; /healthz stays liveness and
probes nothing. Keeping them distinct matters: wiring liveness to dependency
health means an orchestrator restarts KeyCape when a dependency blinks, and a
restart also discards every in-flight login, so the reaction is worse than the
condition it reacts to.
LLDAP is probed with a bind rather than a dial, since a rotated or revoked
service password leaves the port open and every lookup failing -- exactly what
readiness should catch and exactly what a dial would miss. The response names the
failing check but never the reason: the endpoint is unauthenticated and upstream
error text carries hostnames and sometimes credentials-in-URLs. Results are
cached for 2s so an unauthenticated endpoint cannot be used to drive unbounded
upstream traffic, and probes run concurrently under a 3s bound so a hung
dependency makes the endpoint answer rather than hang with it.
SIGTERM and SIGINT now drain in-flight requests for 15s, under the 30s read/write
timeouts so a stuck request cannot outlive the window before SIGKILL.
docs/operations.md states the single-replica topology and why, and three limits
easy to get wrong: the constant key-1 key ID makes same-kid rotation a trap for
consumers caching JWKS, removing a client does not revoke its issued tokens, and
/logout is local only. No throughput figures are given, since nothing here
benchmarks KeyCape. Shared storage and refresh tokens stay excluded, as G08
allows.
Verified in the running executable: 503 naming all three checks failed while
/healthz returned 200, the LLDAP check flipping to ok once started, and 40/40
requests succeeding across a SIGTERM.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NV9oijZukGyGbRQGGKnK4P
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713576@bnt-lap001
Assistant-Session: 384c511d-9bce-4cb8-a676-2aef6c0c8df6
