nginx and certbot were designed assuming they lived on the same host
as the app. They don't - a dedicated proxy Pi terminates TLS and
forwards plain HTTP over the LAN. Kestrel now binds 0.0.0.0 instead
of localhost, the whole certbot procedure moved to the proxy Pi's
side, and pi-main gets a firewall rule restricting the backend ports
to just the proxy Pi's address - otherwise binding all interfaces
would let anything on the LAN skip the proxy's TLS entirely.
New section walks through adding any domain from nothing - plain
HTTP block first (certbot's nginx plugin needs one to validate
against), then the certbot command, then a verification step. The
existing slpsoftware.nl blocks stay as the worked result of that
procedure rather than a separate one-off.
Traced live on the actual Pi: su - gitea-workflow left XDG_RUNTIME_DIR
unset even with the user manager already running and the bus socket
already there. None of su/sudo -i/sudo -u reliably go through
pam_systemd on Debian - only a real SSH login or sshd itself does,
which is what the deploy workflow already uses.
Enabling lingering doesn't start the user manager retroactively, and
sudo -u often drops XDG_RUNTIME_DIR - both produce the same "Failed
to connect to bus" error when setting this up by hand.
The target Pi only has MariaDB, and SQL Server has no ARM64 build at
all - not a config problem, a real gap discovered during deployment
setup. Swapped the EF Core provider, regenerated every migration,
updated connection strings and the backup script everywhere they
appear.
Took two tries to land on a provider that actually works: Pomelo
builds fine against this project's EF Core 10 packages but fails at
runtime (it's compiled against 9's internal API surface, which moved
in 10 wherever Identity/DataProtection force the newer packages).
Oracle's official provider builds and migrates fine but has a real
MariaDB bug in its own migration-lock code, reproduced against a live
database. Kept Oracle's provider and worked around just that one
broken method - everything else it does is correct - rather than
give up more of the stack to chase a workaround.
Verified against a real local MariaDB end to end: all three
migrations applied, both hosts start clean, full suite still green.
AllowedConnectOrigins had a sentry-ingest-origin placeholder that
was simply incorrect - the Sentry tunnel exists precisely so that
exception is never needed. Both SecurityHeaders fields are Umami-only
now, with inline comments on when to set them versus leave them out,
and a note that MasterModule__MasterUrl degrades to a warning rather
than a startup failure when unset.
SlpModularCms is the product; slpsoftware.nl is the customer running
it - the first one, and a fine stand-in for what a real deployment
looks like. Renamed the directories, systemd units, and backup paths
accordingly. Left the backup script's own filename alone since the
already-committed deploy-scp.yaml calls it by that exact name.
Applies to both the app's own release tree and the website upload
path - slpmodularcms/test and slpmodularcms/production side by side
under one parent, not two separately-named directories. Service unit
names stay hyphenated; those aren't folders.
No reason to mint a new system account for every project on the
same Pi - one shared, host-wide account works, namespaced by
directory per project underneath it.
webadmin turned out to be FTP-only and belongs to whoever uploads
customer websites, not to this pipeline. Switched to a separate
deploy account with its own home-directory paths, well away from
where the other sites live, and made the persistent website symlink
cross into webadmin's territory instead of the other way around.
Host setup checklist, real domains and ports, the database backup
script the deploy workflow only ever invokes, and a rollback plan.
Also drafted the FTPS switch procedure for whenever production moves
off the Pi, with a note that shared hosting is likely IIS-based -
that's a bigger change than swapping the transport.
Unit suites already passed individually; this is the full solution
together plus what none of them could check alone: two real hosts
talking to each other, a slave API key encrypted by one process and
decrypted by the next one after a restart, and headers landing on
actual static files instead of a simulated response.
U7 was the last one. State doc corrected: it had U7 down for an
infrastructure design pass that never actually happened, because by
the time units were carved up U7 had shrunk to documentation only.
WEBSITE_WORKSPACE.md, plus the parts of the README that were still
describing the old layout or a manual step the code doesn't need
anymore (the key-ring paragraph, mainly - that one was actively
wrong now, not just stale).
continuous_integration.yaml: six blocking checks, then a separate
publish per environment so a Vite build never gets tagged for the
wrong one, then a call into last commit's deploy workflow. Along the
way: the lint list had drifted (two problems not in the requirement,
one already fixed), and the Umami-origin gate needed a variable pair
of its own since the backend's side of that comparison lives on the
host, not in CI. Pinned the two vulnerable packages while at it.
deploy-scp.yaml: download, back up the production database before
touching anything, upload into a timestamped release, symlink the
persistent website in, switch current and restart, verify /health
with retries, prune old releases only once that check passes. No
sudo, no container actions, one shared interface U5 will call next.
Infrastructure design for U6: single Pi split by directory not port,
systemd --user services instead of sudo, a releases/current/shared
layout that keeps the customer website outside every atomic switch,
and a health check through the existing reverse proxy before pruning
anything. Flags that user-level systemd needs enable-linger or the
service dies the moment the deploy SSH session closes.
Records the stage as complete and logs the two defects that local testing
turned up after U3 and U4 were generated: ciphertext predating the key-ring
move, and an integrity check that could not tell an unreachable slave from
one that does not recognise the master.
Both were fixed in this branch rather than filed. The second is master/slave
domain behaviour rather than deployment work, so the state file says so
plainly — it sits here by decision, not because it belongs to the feature.
Verified at close: build 0 errors, 372 backend tests, 237 frontend tests.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
Starting both hosts after U2 fails to decrypt a stored slave API key: the
row was encrypted under the old file-based key ring, and the database key
ring holds a different, freshly generated key. The key ring working
correctly is exactly why the old key is not found.
U2 moved the ring without a path for ciphertext that predates it. Confirmed
with the user that nothing is deployed yet, so no migration ships and the
first deploy meets an empty database. ASM-08 records the condition and makes
it blocking for U6 if it ever stops holding.
Also notes that the master lets the CryptographicException escape while the
slave catches it and reports "master API key rejected" instead — the
ambiguity U4's event was designed around, showing up on day one.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
U4. Console logging plus Sentry, a same-origin tunnel so ad blockers cannot
silence browser errors, Umami on the admin SPA, and six security events that
alert rules can actually be built on.
The correlation id is the W3C trace id from the ambient Activity, enabled by
one line of ActivityTrackingOptions so every entry from every category carries
it without touching a call site. It propagates across the master/slave
boundary via traceparent, which TraceIdentifier cannot do at all, and it is
the same value ProblemDetails already returns to the browser.
The security events use source-generated LoggerMessage with constant
templates. Sentry groups log events by message, so interpolating an email
address would give every address its own issue and "more than 20 failed
logins in five minutes" could never fire — the events would arrive, be
visible, be tagged, and the alerting would silently be impossible. A test
asserts the rendered message is identical across argument values.
Scrubbing happens in-process, before transmission, and covers Set-Cookie as
well as Cookie: the login response issues the refreshToken there, so
scrubbing only the request side would protect nothing. Transactions are
scrubbed too, because they carry request data and are the channel nobody
thinks of.
The tunnel derives its destination from the DSN once at startup and reads
nothing from the request, which is what separates a tunnel from a
server-side request forgery primitive. Size is capped by a bounded read
rather than by trusting Content-Length, and the endpoint is rate limited.
Two things found along the way. Zod 4's url() hands the value to the URL
constructor, which accepts any scheme — so the existing frontend validation
would have accepted the exact "htp://" typo BR-U4-24 names, and the SPA
would have called a nonexistent origin. Now constrained to http(s). And the
new appsettings comments are verified against the real configuration
provider, because the failure mode if it rejected them is both hosts
refusing to start after a release switch.
One deviation. IAdminTokenValidator was meant to gain a reason-reporting
overload; implemented that way, a substitute returning false by default
silently inverted the access decision while both methods compiled. Two
methods whose difference is invisible at the call site is the defect, so it
is now a single Validate returning AdminTokenResult.
Touches two files from already-committed units: DatabaseMigrationExtensions
(U2) gains a flush before the rethrow, or the one Critical event in the
system dies with the process; AdminTokenValidator (U1) classifies why a
bypass was refused.
Build 0 errors; 366 backend tests pass, up from 315, and 237 frontend tests,
up from 213. tsc clean, eslint clean on every changed file.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
U3. These headers normally come from nginx, but the deployment target does
not allow server configuration, so the application emits them itself. That
changes the failure mode: a bad nginx config fails loudly at reload, while a
middleware that never runs sends nothing and says nothing.
Two policies, defined in code. Strict for /admin, /api/v1 and /health;
relaxed for the public website, which is authored elsewhere by someone who
has never seen this policy. Configuration decides where a policy applies and
which external origins are permitted; it cannot invent a policy that is
subtly permissive.
script-src 'self' under Strict has no 'unsafe-inline' and no 'unsafe-eval',
asserted by a test so that loosening it means deleting a test that says why.
style-src does carry 'unsafe-inline' and cannot not: Radix positions its
overlays with inline style attributes, which nonces cannot reach at all.
Two traps handled explicitly. StartsWithSegments rather than string
StartsWith, because "/administrator".StartsWith("/admin") is true and a
public page would silently lose its inline scripts with no server-side trace.
And all decision logic sits in a static writer rather than in the middleware,
because DefaultHttpContext.Response.OnStarting is a no-op — the obvious
middleware test observes nothing and an assertion that nothing was written
passes for entirely the wrong reason.
An unknown policy name fails ValidateOnStart, so the process exits rather
than quietly serving /admin under the relaxed policy. Origin format is
validated too, beyond what the design asked: a CSP source list silently
ignores a malformed source, so a URL with a path would look configured and
block the script anyway.
BR-U3-22's Umami startup warning is withdrawn (REF-U3-01) — the backend
never sees VITE_UMAMI_WEBSITE_ID. It becomes a blocking CI gate in U5.
Build 0 errors; 315 tests pass, up from 253.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
NFR Design for U3 and U4. Two decisions the earlier stages had deliberately
left open, plus four risks the functional design did not name.
OPEN-01 closed: the correlation ID is the W3C trace ID from the ambient
Activity, with TraceIdentifier as the fallback. It propagates across the
master/slave boundary via traceparent, which TraceIdentifier cannot do at
all, and it is the same value ProblemDetails already returns to the client.
REF-U3-01 raised: BR-U3-22's Umami-origin startup warning cannot work. The
backend never sees VITE_UMAMI_WEBSITE_ID, so the check would either always
warn or never warn. Withdrawn from U3 and replaced by a blocking U5 CI gate
that compares the frontend build variable against that environment's CSP
origins, where both values are visible.
Four additions beyond the functional design:
- Set-Cookie added to the scrub list; the login response issues the
refreshToken there, so scrubbing only the request cookie protects nothing
- SetBeforeSendTransaction alongside SetBeforeSend; transactions carry
request data too
- OnRejected on the rate limiter; today a 429 leaves no trace anywhere
- FlushAsync before the migration-failure rethrow, or the one Critical
event in the system dies with the process
Two traps recorded with tests attached: Sentry groups log events by message
template, so interpolated messages make FR-19's rate-based alert rules
unimplementable while appearing to work; and DefaultHttpContext.Response
.OnStarting is a no-op, so the obvious middleware test asserts nothing.
Three values chosen rather than escalated, each one line to change and all
three listed for review at the end of U4's pattern document: JSON console
outside Development, TracesSampleRate 0.1, tunnel cap 200 KB.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
Records the functional design for the two remaining application units,
before any of their code exists.
Security headers have to come from the application, because relying on
nginx or IIS configuration is exactly what this deployment model rules
out. Strict applies to /admin, /api/v1 and /health; a relaxed policy
applies to the public website, which this repository does not author.
The strict policy needs style-src 'unsafe-inline'. That is not a
shortcut: Radix positions dropdowns and dialogs with inline style
attributes recalculated per click and scroll position, and CSP nonces
apply only to style elements, never to style attributes. No nonce- or
hash-based variant leaves the admin UI working. The exception is bounded
to styles — script-src stays closed, which is where XSS actually lives.
The website's policy is enforcing rather than absent, so every
HTML-serving path carries a CSP and no exception has to be recorded. It
still blocks external script origins, so it remains a real boundary.
HSTS is skipped in development: browsers remember it per host and
localhost is shared with unrelated projects. Every other header applies
locally, so a CSP violation surfaces while developing.
For observability, browser error reports tunnel through the API rather
than going to Sentry directly. Ad blockers block Sentry domains, which
loses errors precisely for the users most likely to have browser
oddities. The tunnel forwards only to the host derived from the
configured DSN — a caller-supplied destination would turn an anonymous
endpoint into a request-forgery primitive.
Two consequences of the chosen options are recorded rather than left
implicit:
Enabling SendDefaultPii attaches request headers, and this application
carries two standing credentials in them. Besides the refreshToken
cookie, X-Master-Api-Key would have been sent to a third party on every
error raised during a master/slave call. The scrub list removes the
whole Cookie header, Authorization, X-Master-Api-Key and the request
body.
Console logging at Information plus structured logging to Sentry would,
taken literally, mean one Sentry event per request — exhausting the free
plan within hours and burying real errors in request noise. The
thresholds are split: console keeps Information, Sentry takes warnings
and above as events with Information as breadcrumbs, so every event
arrives carrying the trail that led to it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
Nothing here is visible in normal operation. Its whole purpose is that
swapping the release directory on deploy cannot silently destroy state.
Data Protection secures the API keys that authenticate master/slave
communication. Two separate defaults would each have destroyed them:
keys are held on the filesystem, which a release swap discards, and the
application discriminator is derived from the content root path, which
changes with every release directory — so even keys stored in a database
would have stopped being derivable. Keys now live in
ApplicationDbContext and the discriminator is a fixed constant.
Losing them produces no error. It produces stored keys that no longer
decrypt, which presents as an apparent network fault between a Master
and its slaves and is easily misdiagnosed. That is also why the tests
assert the resulting configuration rather than the registration: the
XmlRepository must be the EF one and the discriminator must be the
constant, plus a round-trip proving a value encrypted before a deploy is
readable after one. A test that only checked "Data Protection is
registered" would have passed in the broken case too.
Both modules previously called AddDataProtection() themselves. Module
registration runs after the host's, so those calls re-registered the
configuration chain and would have overridden the persistent store while
IDataProtector still resolved. They are removed, with a comment at each
site — the deletion otherwise looks like a regression. Each module's own
test project now guards against it being reintroduced.
ApplicationDbContext also migrates itself at startup. Deploy targets
offer no CLI, so migrations cannot be a manual step on the server.
Failures are classified rather than treated alike: a connection failure
means the database is not up yet, normal when the app and the database
start together after a reboot, and is retried with backoff; a migration
failure means something is broken and fails at once. Either way the
process does not start, which is what makes the liveness health check
trustworthy — an application that cannot reach its schema never answers
/health, so monitoring goes red instead of reporting a healthy instance
that cannot serve a request.
The cost of migrating without a human gate is that migrations must stay
forward-compatible and non-destructive, since rollback is "redeploy the
previous release". The new migration is purely additive.
Also wires this and the preceding hosting commit into both hosts, as
they touch the same lines of Program.cs.
Two constraints are enforced by documentation rather than code, and
belong in the deployment instructions: the key table must never be
pruned, and only one instance may migrate a given database at a time.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
Prepares the single-host layout for deployment. The customer's public
website moves from wwwroot/ to wwwroot/web/, so a CMS deploy can no
longer overwrite content it does not own: with the website in its own
directory, the release directory can be swapped without touching it.
Each front-end gets its own file provider, and both tolerate a missing
directory at startup — a fresh deployment has no website until a
separate workspace deploys one, and the CMS must still serve /admin and
the API. When the website's index.html is absent, an embedded
placeholder is served instead of a 404, which also doubles as proof the
CMS itself is running. The placeholder is embedded in the assembly
rather than shipped into wwwroot/web/, because that directory is owned
and overwritten by the website workspace.
Adds GET /health for uptime monitoring. It reports infrastructure
liveness only and is deliberately NOT the same thing as
/api/v1/Availability/status or /api/v1/System/capabilities: those are
CMS domain state that also serve the master/slave protocol. A healthy
instance can be switched off by design, and a switched-on instance can
be unhealthy, so conflating them would alert on business state and stay
silent on real outages. /health is on the availability gate's bypass
list for the same reason.
Fixes a real defect in the gate's admin bypass. It parsed the bearer
token with ReadJwtToken, which reads claims without verifying the
signature, so an unauthenticated caller could forge an unsigned token
carrying an Owner role claim and bypass the gate that suspends a
customer's site. Protected endpoints still rejected them, so nothing
leaked — but the gate itself was bypassable. The token is now fully
validated against the same parameters as the bearer scheme, resolved
from one shared source so the two cannot drift apart.
Host wiring for these changes lands with the data-durability commit,
since both units touch the same lines of Program.cs.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
Adds the AI-DLC inception record for deploying the CMS as a single .NET
application on hosting where no server configuration is possible.
The reverse-engineering artifacts were regenerated: the previous set
predated the Master module, the Slave host, the solution reorganisation
and single-host serving, all of which matter for deployment. Findings
were verified by running the build, both test suites and the linter
rather than inferred, which surfaced two facts the plan depends on:
the frontend lint gate currently fails (5 errors), and two transitive
packages carry high-severity advisories.
Records 24 functional requirements, 32 traced decisions and a
seven-unit decomposition whose ordering is load-bearing: durability
work must land before the first automated deploy, or the very first
deploy is the one that silently breaks master/slave trust.
Two conflicts found while designing and carried into the units:
- Both modules call AddDataProtection(), which runs after the host and
would override a persistent key store while still passing any
registration test.
- The availability gate runs before authentication, so its admin
bypass cannot read HttpContext.User.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HHoJpxYXzHACSQguHrC5fw
Marks aidlc-state.md as closed and logs the commit/push handoff in
audit.md; these notes were written after the prior commit had already
been created.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Frontend (Unit 2 completion): dual dev-server tooling (pnpm dev:slave,
pnpm dev:all), per-instance browser tab titles, and a backend
capability check (SystemController + useSystemCapabilities +
ModuleGuard) so a Master-only page is hidden on a slave instance
instead of assuming every backend has every module.
Master/slave protocol fixes surfaced by actually running master and
slave side by side locally:
- Deactivating a CMS instance (Inactive) now releases the slave's
master gate instead of leaving it stuck on its last pushed status.
- The periodic integrity check now also re-pushes status to every
reachable slave (previously URL-verification only) and runs once
immediately on startup.
- Added the originally-specified (but never implemented) slave-pull
path: a slave now periodically polls its own status from the master
(GET /api/v1/SlaveStatus) and fails open to Available if the master
is unreachable for too long, complementing the existing push.
- The slave's own Settings page can no longer "successfully" change
local availability while the master controls it; it's now locked
with an explanatory banner and the backend rejects the write with
409 instead of silently no-op'ing it.
- CMS instance status badges now match the dashboard's color/icon
styling instead of a plain grey badge.
Also corrected the master-cms-module design docs to match this
as-built behavior, and flagged (without a full rewrite) a larger,
pre-existing divergence between its inception-stage application
design and what construction actually built.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Relocates ModuleOrchestrator, ServiceCollectionExtensions, and
ApiPrefixConvention from SlpModularCms.Api into SlpModularCms.Core.Hosting
so a new Master-less SlpModularCms.Api.Slave host project (ports 5285/7222)
can share the same bootstrap code without duplicating it. This lets a
developer run a master instance and a slave instance side by side locally
to test the master/slave connection, without touching the existing
master/slave protocol itself.
Relocates the two orchestrator/convention test files from
Modules.Identity.Tests to Core.Tests, dropping an incidental
ProjectReference to SlpModularCms.Api that existed only for those tests.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Finishes the master-cms-module feature (Units 1-4): runs Build and
Test across master-backend, slave-availability-extension and
frontend-cms-page, fixes a missing Availability EF migration for
MasterRegistration and a TanStack Query v5 mutation-callback type
break, adds the missing MasterModule appsettings section, and
documents the module in README.md. Also seeds a tech-debt-backlog
feature to track dead config and pre-existing/introduced frontend
lint findings for later cleanup.
- Add @tanstack/react-query 5.101.0; wrap app with QueryClientProvider
- Add AvailabilityStatus type and AvailabilityResponse to api/types.ts
- Implement useAvailabilityStatus (staleTime 30s, stale-on-error preserved)
- Add AvailabilityStatusBadge with green/amber/red states and stale indicator
- Replace DashboardPage placeholder card with live availability widget
- Add MSW availability handler; update test/utils with QueryClientProvider
- 55/55 tests pass (FR-05)
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
Functional design documents were written based on self-answered questions
instead of asking the user. Removed pre-written docs and restored the plan
with open questions for the user to answer.
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
Sluijsens
