Bump the OpenCode V2 binary in the devenv image from 2.0.12 to the
current npm latest tag (2.0.15) and refresh the per-architecture
SHA-256 checksums for both arm64 and amd64 tarballs.
AI-assisted-by: deepseek-v4.1-flash
A compose port mapping delivers traffic to the container address,
never to loopback, so `listen 127.0.0.1:8082` made
`ports: <host>:8082` fail from the host. Both configs (image
template and devenv) now use `listen 8082`, which binds every
interface and matches the implicit bind of the public
`listen 8080 default_server`.
Rewrite both block comments to state the new bind and who decides
access from outside the host. The scrape URI stays on
127.0.0.1:8082: it still reaches the socket.
AI-assisted-by: mimo-v2.6-flash-free
* 🐳 Add upstream diagnostics to nginx access log
Enrich every access-log line with the internal journey of the request:
the status the backend answered (us), the time spent connecting to it
(uct), the time spent waiting for its answer (urt) and the internal
address that served the request (ua).
A plain 502 line used to say nothing about where the request died. With
this format, the tail of the line classifies the failure: connection
rejected, backend accepted and hung (uct + urt under 1s), or backend
stuck until read timeout. This was the missing witness in the Sep 20
incident, where nginx received connection resets with zero timeouts and
zero rejections.
Applied both to the production image template and the devenv config.
With proxy_pass on variables there is no upstream keepalive, so uct
measures one real TCP connection per request.
Parsing the new fields (us, uct, urt, ua) on the log shipper is left to
ops, so they can be filtered in Loki.
AI-assisted-by: glm-5.3-flash
* 🐳 Add stub_status endpoint for nginx metrics
Add a dedicated localhost-only server (listen 127.0.0.1:8082) exposing
/stub_status next to every other location of the public server. Ops can
run the official nginx-prometheus-exporter as a sidecar against
http://127.0.0.1:8082/stub_status and get nginx_connections_active,
accepted vs handled, reading/writing/waiting and request rates in
Prometheus.
Binding it to localhost and its own server keeps it unreachable from
outside the host and out of the public surface, and access_log off
avoids polluting Loki with one line per Prometheus scrape. The base
image already ships stub_status compiled in, so no image rebuild is
needed.
Applied both to the production image template and the devenv config.
AI-assisted-by: glm-5.3-flash
* ✨ Expose http server gate metrics (worker and connector)
The backend already measured dispatch latency but nothing reported the
state of the "house door": the xnio worker queue and threads, and the
monitor-level listener counters. This was the exact blind spot of the
Sep 20 incident, where the server kept answering health checks while it
accepted connections and dropped them without response.
Add a periodic metrics sampler that lives and dies with the http
server (single daemon thread, 15s interval, each sample guarded so an
unexpected error does not cancel subsequent runs) and publishes:
- worker (xnio MXBean gauges): penpot_http_worker_queue_size,
busy_threads, pool_size and max_pool_size. Negative samples are
discarded: the MXBean transiently reports -1 on the busy thread
count (verified live), and a stale negative would read as zero.
- listener (Undertow connector statistics, enabled via the new
:server/statistics yetti option): penpot_http_connector_active*
_connections gauge and requests_total / errors_total counters.
Undertow exposes absolute totals, so the sampler keeps a watermark
atom and publishes deltas, skipping (and moving forward past) a
counter reset.
The connector-level part depends on yetti v11.11, which now accepts
a :server/statistics server option (patch authored and released
upstream; before it, ListenerInfo#getConnectorStatistics always
returned nil).
New tests cover the samplers with fake MXBean/collector statistics
against real prometheus collectors, including the negative-sample
filter, the delta/watermark logic and the sampler lifecycle.
AI-assisted-by: glm-5.3-flash
* 🐛 Include jdk.management in the backend runtime JRE
The production image builds a trimmed JRE with jlink and omitted
jdk.management. Without that module the OS MXBean is
sun.management.BaseOperatingSystemImpl, which has no
getProcessCpuTime, getOpenFileDescriptorCount nor
getMaxFileDescriptorCount. The prometheus client StandardExports
reads those getters reflectively and collect() swallows the
NoSuchMethodException, so process_open_fds, process_max_fds and
process_cpu_seconds_total silently disappeared from /metrics while
the other process_* families kept flowing.
Verified against Prometheus: the app job only ever exposed
process_start_time_seconds, process_virtual_memory_bytes and
process_resident_memory_bytes; the fd and cpu families were absent.
Reproduced locally by running the backend metrics registry on a JRE
built with the same jlink module list (false/false/false) and on one
with jdk.management added (true/true/true).
Add the module to --add-modules and pin the metric contract with
backend-tests.metrics-test.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Build the http metrics sampler on promesa.exec
Replace the hand-rolled ScheduledThreadPoolExecutor and ThreadFactory
with promesa.exec primitives: px/scheduled-executor with a daemon
thread factory, and a px/schedule chain that reschedules the next
sample when the current one finishes.
Beyond fitting the existing periodic-task pattern (worker/cron,
rpc/rlimit), the chained schedule makes the docstring promise real:
with scheduleAtFixedRate an exception escaping the runnable cancelled
the following executions, while the reschedule now happens in a
finally block.
The sampler shutdown uses px/shutdown-now (shutdown! is deprecated in
promesa 12.0.0) to cancel the pending sample, keeping the previous
halt semantics.
The lifecycle test moves to the promesa predicates and a new test
covers the error-resilience promise: the first sample runs, throws,
and the next one is still scheduled.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Tighten the http metrics samplers
The samplers are leaf functions: they receive what they need and
publish it. Drop the internal nil guards (if there is no metrics
instance or no mxbean there is nothing to call them for) and move the
checks to the boundary, where the optional data is resolved:
sample-http-metrics now short-circuits with some-> and when-let.
Write the four worker gauges as four static operations instead of a
vector of pairs walked by doseq: the set is fixed, so the collection
only adds an allocation and hides each operation.
Drop the ! suffix from the sample-*-metrics family: ! marks a function
whose contract is to mutate state, while these report, and the mutation
happens in the mtx/run! they call. The constant true return, which only
existed so the removed guard tests could assert it, goes away too.
Tests follow the move: the internal-guard tests are replaced by one
boundary test (a nil server publishes nothing).
AI-assisted-by: deepseek-v4.1-flash
* 📚 Add the function design rules memory
Document the rules that came out of the http metrics sampler review:
preconditions are checked at the boundary instead of re-checked in the
core, optional-by-design data is guarded where the optionality is born,
a fixed set of operations is written statically, ! marks mutation and
not reporting, and production code is not shaped for tests.
Also state in the memory maintenance guide that memories must not use
manual line wrapping.
Linked from critical-info so it is read when designing a solution or
an API, not only when touching the samplers.
AI-assisted-by: deepseek-v4.1-flash
* 📚 Unwrap the critical-info memory lines
The memory maintenance guide forbids manual line wrapping, so rewrite
critical-info with one line per bullet and paragraph. A stray `*` at
the start of one continuation line is dropped.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Drop the redundant guard in the http server halt
create-metrics-sampler always returns the scheduler, so the sampler is
always present when integrant calls halt-key!; the nil check was dead
code, same as the yt/stop! call next to it.
AI-assisted-by: deepseek-v4.1-flash
* ✨ Add srepl helper to delete profiles by email
Add `delete-profiles-by-email!` to app.srepl.main. It accepts a
single email, a comma separated list of emails or a coll of emails,
resolves each profile, logs it to audit and enqueues the
delete-object task. The deleted-at is backdated with the configured
deletion-delay so profiles and their owned teams are purged on the
next gc pass.
Extract the per-email deletion logic into a private fn and reuse it
from `delete-profiles-in-bulk!`. Add tests for the new
`parse-emails` helper.
AI-assisted-by: glm-5.3-flash
* ⬆️ Upgrade MCP SDK to v2 and remove HTTP sessions
MCP's per-request protocol removes the need to retain HTTP sessions.
Use the v2 handler to manage each request's transport and lifecycle,
so requests can reach any server instance without session affinity
or the workaround that adopts sessions through private SDK fields.
Keep legacy SSE support and the shared plugin and Redis bridges.
Remove the shared expiry checker, including legacy SSE idle expiry;
SSE connections now remain until disconnection or server shutdown.
Verify stateless requests, token isolation, and legacy compatibility.
Resolves#11827
AI-assisted-by: gpt-6-astra
* 🔥 Remove legacy MCP SSE support
Use Streamable HTTP as the sole MCP client transport so the server no
longer needs a separate SSE connection registry or lifecycle.
Remove /sse and /messages, their nginx routes, and the server-legacy
dependency. Legacy SSE clients must switch to /mcp; older Streamable
HTTP clients remain supported. Document the migration and verify that
the removed endpoints return 404.
Resolves#11846
AI-assisted-by: gpt-6-astra
A compose port mapping delivers traffic to the container address,
never to loopback, so `listen 127.0.0.1:8082` made
`ports: <host>:8082` fail from the host. Both configs (image
template and devenv) now use `listen 8082`, which binds every
interface and matches the implicit bind of the public
`listen 8080 default_server`.
Rewrite both block comments to state the new bind and who decides
access from outside the host. The scrape URI stays on
127.0.0.1:8082: it still reaches the socket.
AI-assisted-by: mimo-v2.6-flash-free
* 🐳 Add upstream diagnostics to nginx access log
Enrich every access-log line with the internal journey of the request:
the status the backend answered (us), the time spent connecting to it
(uct), the time spent waiting for its answer (urt) and the internal
address that served the request (ua).
A plain 502 line used to say nothing about where the request died. With
this format, the tail of the line classifies the failure: connection
rejected, backend accepted and hung (uct + urt under 1s), or backend
stuck until read timeout. This was the missing witness in the Sep 20
incident, where nginx received connection resets with zero timeouts and
zero rejections.
Applied both to the production image template and the devenv config.
With proxy_pass on variables there is no upstream keepalive, so uct
measures one real TCP connection per request.
Parsing the new fields (us, uct, urt, ua) on the log shipper is left to
ops, so they can be filtered in Loki.
AI-assisted-by: glm-5.3-flash
* 🐳 Add stub_status endpoint for nginx metrics
Add a dedicated localhost-only server (listen 127.0.0.1:8082) exposing
/stub_status next to every other location of the public server. Ops can
run the official nginx-prometheus-exporter as a sidecar against
http://127.0.0.1:8082/stub_status and get nginx_connections_active,
accepted vs handled, reading/writing/waiting and request rates in
Prometheus.
Binding it to localhost and its own server keeps it unreachable from
outside the host and out of the public surface, and access_log off
avoids polluting Loki with one line per Prometheus scrape. The base
image already ships stub_status compiled in, so no image rebuild is
needed.
Applied both to the production image template and the devenv config.
AI-assisted-by: glm-5.3-flash
* ✨ Expose http server gate metrics (worker and connector)
The backend already measured dispatch latency but nothing reported the
state of the "house door": the xnio worker queue and threads, and the
monitor-level listener counters. This was the exact blind spot of the
Sep 20 incident, where the server kept answering health checks while it
accepted connections and dropped them without response.
Add a periodic metrics sampler that lives and dies with the http
server (single daemon thread, 15s interval, each sample guarded so an
unexpected error does not cancel subsequent runs) and publishes:
- worker (xnio MXBean gauges): penpot_http_worker_queue_size,
busy_threads, pool_size and max_pool_size. Negative samples are
discarded: the MXBean transiently reports -1 on the busy thread
count (verified live), and a stale negative would read as zero.
- listener (Undertow connector statistics, enabled via the new
:server/statistics yetti option): penpot_http_connector_active*
_connections gauge and requests_total / errors_total counters.
Undertow exposes absolute totals, so the sampler keeps a watermark
atom and publishes deltas, skipping (and moving forward past) a
counter reset.
The connector-level part depends on yetti v11.11, which now accepts
a :server/statistics server option (patch authored and released
upstream; before it, ListenerInfo#getConnectorStatistics always
returned nil).
New tests cover the samplers with fake MXBean/collector statistics
against real prometheus collectors, including the negative-sample
filter, the delta/watermark logic and the sampler lifecycle.
AI-assisted-by: glm-5.3-flash
* 🐛 Include jdk.management in the backend runtime JRE
The production image builds a trimmed JRE with jlink and omitted
jdk.management. Without that module the OS MXBean is
sun.management.BaseOperatingSystemImpl, which has no
getProcessCpuTime, getOpenFileDescriptorCount nor
getMaxFileDescriptorCount. The prometheus client StandardExports
reads those getters reflectively and collect() swallows the
NoSuchMethodException, so process_open_fds, process_max_fds and
process_cpu_seconds_total silently disappeared from /metrics while
the other process_* families kept flowing.
Verified against Prometheus: the app job only ever exposed
process_start_time_seconds, process_virtual_memory_bytes and
process_resident_memory_bytes; the fd and cpu families were absent.
Reproduced locally by running the backend metrics registry on a JRE
built with the same jlink module list (false/false/false) and on one
with jdk.management added (true/true/true).
Add the module to --add-modules and pin the metric contract with
backend-tests.metrics-test.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Build the http metrics sampler on promesa.exec
Replace the hand-rolled ScheduledThreadPoolExecutor and ThreadFactory
with promesa.exec primitives: px/scheduled-executor with a daemon
thread factory, and a px/schedule chain that reschedules the next
sample when the current one finishes.
Beyond fitting the existing periodic-task pattern (worker/cron,
rpc/rlimit), the chained schedule makes the docstring promise real:
with scheduleAtFixedRate an exception escaping the runnable cancelled
the following executions, while the reschedule now happens in a
finally block.
The sampler shutdown uses px/shutdown-now (shutdown! is deprecated in
promesa 12.0.0) to cancel the pending sample, keeping the previous
halt semantics.
The lifecycle test moves to the promesa predicates and a new test
covers the error-resilience promise: the first sample runs, throws,
and the next one is still scheduled.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Tighten the http metrics samplers
The samplers are leaf functions: they receive what they need and
publish it. Drop the internal nil guards (if there is no metrics
instance or no mxbean there is nothing to call them for) and move the
checks to the boundary, where the optional data is resolved:
sample-http-metrics now short-circuits with some-> and when-let.
Write the four worker gauges as four static operations instead of a
vector of pairs walked by doseq: the set is fixed, so the collection
only adds an allocation and hides each operation.
Drop the ! suffix from the sample-*-metrics family: ! marks a function
whose contract is to mutate state, while these report, and the mutation
happens in the mtx/run! they call. The constant true return, which only
existed so the removed guard tests could assert it, goes away too.
Tests follow the move: the internal-guard tests are replaced by one
boundary test (a nil server publishes nothing).
AI-assisted-by: deepseek-v4.1-flash
* 📚 Add the function design rules memory
Document the rules that came out of the http metrics sampler review:
preconditions are checked at the boundary instead of re-checked in the
core, optional-by-design data is guarded where the optionality is born,
a fixed set of operations is written statically, ! marks mutation and
not reporting, and production code is not shaped for tests.
Also state in the memory maintenance guide that memories must not use
manual line wrapping.
Linked from critical-info so it is read when designing a solution or
an API, not only when touching the samplers.
AI-assisted-by: deepseek-v4.1-flash
* 📚 Unwrap the critical-info memory lines
The memory maintenance guide forbids manual line wrapping, so rewrite
critical-info with one line per bullet and paragraph. A stray `*` at
the start of one continuation line is dropped.
AI-assisted-by: deepseek-v4.1-flash
* ♻️ Drop the redundant guard in the http server halt
create-metrics-sampler always returns the scheduler, so the sampler is
always present when integrant calls halt-key!; the nil check was dead
code, same as the yt/stop! call next to it.
AI-assisted-by: deepseek-v4.1-flash
* ✨ Add srepl helper to delete profiles by email
Add `delete-profiles-by-email!` to app.srepl.main. It accepts a
single email, a comma separated list of emails or a coll of emails,
resolves each profile, logs it to audit and enqueues the
delete-object task. The deleted-at is backdated with the configured
deletion-delay so profiles and their owned teams are purged on the
next gc pass.
Extract the per-email deletion logic into a private fn and reuse it
from `delete-profiles-in-bulk!`. Add tests for the new
`parse-emails` helper.
AI-assisted-by: glm-5.3-flash
Migrate docker/images/Dockerfile.media-processor from ubuntu:26.04 to
dhi.io/node (Debian 13/trixie), which also drops the manual Node tarball
download since the base image ships it. The -dev tag stays as the final
image: fontforge, woff2 and the graphics libraries are needed at runtime.
scripts/build now assembles the release bundle under target/ (dist/ plus
the manifests and a generated setup script), the way the other modules do,
since esbuild leaves the runtime dependencies external. manage.sh gains
build-media-processor-bundle and build-media-processor-docker-image, both
wired into build-bundle and build-docker-images.
The CI workflows are intentionally left untouched: the module is still work
in progress and its images are not published yet, so this only enables
local builds."
Signed-off-by: David Barragán Merino <david.barragan@kaleidos.net>
dhi.io/node:24.20.0-debian13-dev does not pre-create /usr/local/bin
the way a regular Debian image does, so tar -xzf ... -C /usr/local/bin
failed with "Cannot open: No such file or directory" right after the
checksum check passed. Introduced in #11790 when pnpm moved from
Corepack to a downloaded standalone binary.
Dockerfile.media-processor and docker/devenv/Dockerfile are unaffected:
both extract into /opt/node/bin, which already exists from the prior
Node.js install step.
Signed-off-by: David Barragán Merino <david.barragan@kaleidos.net>
Mirrors the devenv change in 3385a65 (docker/devenv/docker-compose.infra.yml).
Mailpit keeps messages in memory only (no persistence), consistent with
mailcatcher's previous behavior; the volume and MP_DATABASE env var are
left commented for anyone who wants to opt in later. UI port changed from
1080 to Mailpit's native 8025.
Signed-off-by: David Barragán Merino <david.barragan@kaleidos.net>
dhi.io/node:24.20.0-debian13-dev does not pre-create /usr/local/bin
the way a regular Debian image does, so tar -xzf ... -C /usr/local/bin
failed with "Cannot open: No such file or directory" right after the
checksum check passed. Introduced in #11790 when pnpm moved from
Corepack to a downloaded standalone binary.
Dockerfile.media-processor and docker/devenv/Dockerfile are unaffected:
both extract into /opt/node/bin, which already exists from the prior
Node.js install step.
Signed-off-by: David Barragán Merino <david.barragan@kaleidos.net>
* 🌐 Complete Catalan translations in frontend
Complete the Catalan (ca.po) locale to 100% coverage against en.po,
using es.po as support reference. Adds the 1439 missing entries
across workspace, dashboard, labels, shortcuts, subscription,
errors, modals and onboarding, keeping vosaltres treatment and
IEC/Termcat terminology consistent with the existing strings.
Normalizes placeholders and plural forms, drops the 14 stale
obsolete entries and canonicalizes the file with the repo
translations script.
Closes#11739
AI-assisted-by: muse-spark-1.3-contributor
* 📚 Add frontend translations memory with Catalan criteria
Record the PO workflow, the sync fuzzy-flag gotcha and the
Catalan glossary and tone agreed upon while completing ca.po,
and link the new memory from the frontend core routing.
AI-assisted-by: muse-spark-1.3-contributor
* 🔧 Add gettext to devenv image
Provide msgfmt and msgattrib in the dev environment for
checking PO translation files.
AI-assisted-by: muse-spark-1.3-contributor
* 🌐 Fix Catalan translations and add PO checker
Review of the missing-whitespace pattern found ~90 glued words
across 75 entries, plus 4 lost plural forms and 2 placeholder
mismatches verified against tr call sites. All fixed in ca.po.
Adds frontend/scripts/check-translations.js (vocabulary-free PO
QA: glued words, punctuation, placeholders, plurals) with
--self-test, wired as pnpm run check-translations and
documented in mem:frontend/translations.
AI-assisted-by: muse-spark-1.3-contributor
* 🌐 Multi-locale PO checker with word catalogs
Split the checker engine from its word lists: ca/es catalogs now
live in scripts/check-translations/words.<locale>.txt and all
messages are in English. Adds an es seed (calibrated to zero
errors) and fixes 7 typos it found in es.po. Universal checks
(placeholders, plurals, punctuation) run without a catalog.
AI-assisted-by: muse-spark-1.3-contributor
* 🌐 Merge PO checker into translations.js
Fold check-translations.js into translations.js as a check
subcommand reusing its locale helpers; word lists stay in
scripts/check-translations/words.<locale>.txt. Also fixes the
getopts stopEarly bug that made -l useless after the command
(sync -l ca synced every locale), drops dead lodash import
and code, unifies help and exit codes. Removes the
check-translations package alias; use translations.js
check -l <locale> with explicit -l.
AI-assisted-by: muse-spark-1.3-contributor
* 🌐 Keep unused placeholders out of the gate
Reverts the %s-stripping on unused auth.terms-privacy-agreement:
the links mirror its markdown sibling and a reactivation may
need them. Placeholder mismatches on #, unused keys now warn
instead of failing, and the rule is recorded in
mem:frontend/translations.
AI-assisted-by: muse-spark-1.3-contributor
* ⬆️ Update devenv dependencies
Update Node.js, OpenCode, clj-kondo, Babashka, Pixi, GitHub CLI, uv,
and Serena to their current stable releases.
AI-assisted-by: gpt-5.6-sol
* ⬆️ Update devenv to Java 27
Use Zulu JDK 27 in the development image for compatibility testing.
Update the official checksums for both supported architectures.
AI-assisted-by: gpt-5.6-sol
* 🐳 Replace MinIO with RustFS in devenv
Run RustFS as the development S3 service and wait for its health check.
Install a pinned AWS CLI with checksums and use it to create the bucket
idempotently from each backend entry point.
Keep the old MinIO volume untouched and use a new RustFS volume.
AI-assisted-by: gpt-5.6-sol
* 🐳 Replace MailCatcher with persistent Mailpit
Run Mailpit as the devenv SMTP sink while preserving mailer:1025 and the
localhost:1080 UI.
Store its SQLite inbox in a named volume and wait for the readiness
endpoint before starting runtime containers. Bind the web UI to loopback so
development emails stay local.
AI-assisted-by: gpt-5.6-sol
* ⬆️ Update Node.js to 24.21.0
Align the host NVM version with the Node.js version used by devenv.
AI-assisted-by: gpt-5.6-sol
* ⬆️ Update devenv to PostgreSQL 18.6
Run PostgreSQL 18 with its versioned volume layout and a TCP readiness
check that ignores the temporary initialization server.
Install the matching client, create penpot_nexus, and preserve the old
PostgreSQL 16 volume for rollback or logical migration.
AI-assisted-by: gpt-5.6-sol
* 🐳 Expose RustFS ports in devenv
Publish the RustFS S3 API and management console on localhost port 9000
and 9001.
Keep both bindings on loopback so object storage is not exposed to the local
network.
AI-assisted-by: gpt-5.6-sol
* 🐳 Install standalone pnpm in devenv
Install pnpm 12.5.0 from architecture-specific release archives and
verify their published checksums.
Remove the Corepack setup while allowing pnpm to honor the project
packageManager pins.
AI-assisted-by: gpt-5.6-sol
* 🔥 Remove corepack, use system pnpm everywhere
Corepack is gone from Node 25+, so every `corepack enable` call
fails. pnpm now ships as a system binary (devenv, CI runners and
Docker images install it directly) and auto-downloads the version
pinned in `packageManager` on mismatch.
Scripts, workflows and Dockerfiles call `pnpm` straight away; the
three deploy workflows use a single `pnpm/setup@v2` step; and the
new `scripts/sync-pnpm-version` stamps all 35 `packageManager`
fields from the system pnpm, replacing the `corepack use` sweep.
AI-assisted-by: muse-spark-1.3-contributor
* 🐛 Fix exporter watch missing render-wasm build step
The exporter watch compiled CLJS requiring the generated
src/app/wasm/shared.js, which only render-wasm/build export
produces. Without it shadow-cljs failed with a cryptic missing
./shared.js dependency. Run build:wasm before watching, as
the frontend watch:app and exporter scripts/build already do.
AI-assisted-by: muse-spark-1.3-contributor
* 🔧 Add opencode V2 support and adapt plugins
Register the penpot tools for both opencode V1 (server())
and V2 (setup() with JSON Schema inputs) from a single
dependency-free plugin file, sharing the psql and
paren-repair runners between both paths.
Install the opencode2 binary side-by-side with V1 in the
devenv image and document the dual registration in the
paren-repair and psql memories.
AI-assisted-by: muse-spark-1.3-contributor
* ⬆️ Update pnpm and opencode
The frontend build now emits the sha256 hashes of the inline scripts of every page it writes into resources/public, the image moves them out of the document root, and the entrypoint splices them into the default script-src. This removes one of the two reasons why enforcing mode was not usable.
The hashes are computed on the rendered output rather than on the mustache templates, since the digest covers the exact bytes served between the script tags. All four served pages contribute, not just index.html: challenge.html handles the redirect, render.html is loaded by the exporter in a headless browser, and rasterizer.html is initialised by the frontend itself, so leaving any of them out would have broken those paths under enforcing mode. The storybook previews are excluded because that container does not serve them.
A bundle predating this change yields no hashes and the policy stays as it was, so older bundles keep building.
The three external locations were also passing through the security headers of their upstreams. raw.githubusercontent.com returns its own Content-Security-Policy and both it and fonts.googleapis.com return Strict-Transport-Security. Browsers enforce the intersection of every policy they receive, so the upstream one takes precedence on those responses, and the HSTS one lands on our own host, meaning a deployment that deliberately disables HSTS would get it set anyway by a third party. Hide all three at the proxy.
Signed-off-by: David Barragán Merino <david.barragan@kaleidos.net>
The three locations of nginx-external-locations.conf define their own add_header directives, which under nginx's inheritance rules discards every add_header from the enclosing server block. They were therefore served without any of the four security headers already shipped, and would equally have been served without the new CSP and HSTS ones.
Include the common file in all three, which is the pattern the development environment already follows for the equivalent locations.
Signed-off-by: David Barragán Merino <david.barragan@kaleidos.net>
Ship both headers from the image so that every deployment starts from a sensible default instead of each installation deriving its own policy. Report-only mode never blocks a request, so this changes no behaviour for existing deployments, and HSTS stays absent unless PENPOT_PUBLIC_URI declares an https scheme.
The policy can be narrow because the frontend already reverse proxies its own external dependencies, so 'self' covers them. What it must permit beyond that comes from the code: 'wasm-unsafe-eval' for the render engine, 'unsafe-inline' styles for the inline style attributes of the UI, and blob:/data: for thumbnails, exports and fonts.
Closes#11374
Signed-off-by: David Barragán Merino <david.barragan@kaleidos.net>
AI-assisted-by: Claude
* ⬆️ Update pnpm and its deps
* ⬆️ Update JVM dependencies in backend and common
Update several JVM dependencies across backend and common:
- passay 1.6.6 -> 2.0.0 (package reorg, ctor-based rules)
- siphash 2.0.0 -> 3.0.0 (SipHasher* renamed to SipHash*)
- lettuce-core, guava, sqlite-jdbc, jsoup, lz4-java, markdown-clj,
awssdk s3/sts, selmer, jackson-core/databind, shadow-cljs
Adapt passay validation to the new API (moved packages, constructor
configuration) and siphash to the renamed classes. Add tests for
password validation and UUID advisory-lock hashing.
AI-assisted-by: deepseek-v4-flash
* ⬆️ Update node on docker images
* 📎 Minor fixes related to pnpm12 compatibility
* ✨ Add export job model, store and scheduler to exporter
* ✨ Render wasm exports on pooled worker threads
* ✨ Add export job REST API to exporter
* ✨ Use export job API and allow cancelling wasm exports
* 🔧 Show export jobs in the internal debug panel
* 🔧 Pass flags and export job settings to the exporter container
* 📚 Document the exporter job API and its redis layout
On self hosted installs /js/config.js is regenerated from PENPOT_FLAGS
on every container start, but nginx served it with the same
`public, max-age=604800` used for build assets, and index.html versions
it only by the build. A flags only change therefore leaves the URL
untouched, so a browser that had already loaded the app kept using its
cached copy for up to a week: enabling a flag such as
enable-login-with-google had no visible effect for returning users
until the cache expired or they cleared their site data.
Serve that one file with the same no-store headers already used for
index.html, which is the other file whose contents change without its
URL changing. Every other static asset keeps the long lived cache.
Fixes#10556.
Signed-off-by: Filip Sajdak <filip.sajdak@siili.com>
Co-authored-by: Andrey Antukh <niwi@niwi.nz>
The global `proxy_set_header Host $http_host;` forwarded the client-facing
Host to internal proxy_pass calls (backend/exporter), breaking mTLS routing
in service-mesh setups (e.g. Istio STRICT mode), which match outbound
requests to a cluster based on Host/:authority.
Explicitly set `Host $proxy_host` on /api, /assets, /api/export, /readyz
and /ws/notifications so these calls always target the correct internal
service host, independent of the client's original Host header.
Fixes#10835
Signed-off-by: Sebastien MALOT <sebastien.malot@pm.gouv.fr>
Co-authored-by: Sebastien MALOT <sebastien.malot@pm.gouv.fr>
The generated /etc/nginx/overrides/server.d/mcp-locations.conf used a
plain proxy_pass target (e.g. `proxy_pass http://penpot-mcp:4402;`)
where $PENPOT_MCP_URI/$PENPOT_MCP_URI_WS are shell variables substituted
once by envsubst in nginx-entrypoint.sh at container startup, not nginx
variables. nginx resolves a literal proxy_pass hostname once when the
config loads and never re-checks it, so the existing
`resolver 127.0.0.11 valid=10s;` directive in
overrides/http.d/resolvers.conf has no effect on these three locations
- it only applies to nginx variables evaluated per-request.
In multi-container deployments where the penpot-mcp container restarts
or is recreated independently of penpot-frontend (image update, OOM,
orchestrator reschedule), it gets a new IP from Docker's/the
orchestrator's DNS, and the frontend's nginx keeps forwarding to the
old, now-dead address until penpot-frontend itself is restarted. This
surfaces to users as `wss://<host>/mcp/ws` failing to connect from the
browser after enabling the MCP plugin, with
`connect() failed (111: Connection refused)` in the frontend's nginx
logs.
Route each location through a `set $var ...; proxy_pass $var;` pair so
proxy_pass evaluates a real nginx variable, letting the pre-existing
resolver directive re-resolve penpot-mcp within its 10s TTL instead of
caching the address for the container's lifetime.
For /mcp/stream and /mcp/sse, the set value also appends
$is_args$args explicitly: when proxy_pass targets a variable AND that
variable's value includes a URI/path component, nginx does not
automatically forward the original request's query string the way it
does for a static proxy_pass target - it must be appended by hand, or
the userToken query parameter used for multi-user authentication is
silently dropped before reaching the MCP server. /mcp/ws has no path
component in its target so it isn't affected by this and needed no
such change.
Verified locally: force-recreated the penpot-mcp container onto a
different IP while leaving penpot-frontend untouched; the /mcp/ws
WebSocket upgrade kept returning 101 Switching Protocols throughout,
both immediately and after the resolver's TTL window. Separately
verified /mcp/stream: a POST with ?userToken=... now shows up
server-side as userTokenFp=<redacted first 8 chars> instead of <none>,
and an actual MCP client (Claude Code) using this proxy can now call
authenticated tools like execute_code successfully.
Signed-off-by: Jules LaPrairie <jules@lucidbox.ca>
* ✨ Add media-processor service for image and font processing
Externalizes ImageMagick and FontForge subprocess invocations into a
separate Node.js HTTP service (media-processor/). Backend dispatches
via feature flag :use-remote-media-processing.
Key changes:
- media-processor module (TypeScript, Express 5, Sharp, FontForge/woff)
- POST /api/image/info, /api/image/thumbnail, /api/font/generate
- Resource limits: 128MP rejection, prlimit (512MB + 30s CPU)
- Streaming multipart via SequenceInputStream
- app.media split into validation (leaf), local (shell impls), remote (HTTP)
- Schema enforcement: :upload and :input schemas in validation namespace
- Configurable timeout (PENPOT_MEDIA_PROCESSING_SERVICE_TIMEOUT)
- 78 tests across 4 files (image, font, middleware, config)
- FontForge path escaping for command injection prevention
- Parallel font variant conversions with Promise.all
AI-assisted-by: mimo-v2.5-pro
* 🐳 Revert docker-compose changes from media-processor commit
Remove docker-compose.yaml modifications that were part of the media-processor
service commit. The media-processor service definition, flags, and environment
variables are reverted to their previous state.
AI-assisted-by: qwen3.7-plus
* ⬆️ Update dependencies
* 🐛 Fix PR review issues in media-processor
- Font path bug: sfntToWoff and woff2ToSfnt now copy input to temp dir
when input is a file path, ensuring output lands in expected location
- Error preservation: execCommand preserves killed/signal/code properties
from child process errors for OOM detection
- Content-Length: service-multipart-request calculates and includes
Content-Length header for streaming multipart requests
AI-assisted-by: qwen3.7-plus
* 🐛 Fix code review issues in media-processor
- Rename PENPOT_MEDIA_PROCESSOR_SECRET_KEY to PENPOT_MEDIA_PROCESSOR_SHARED_KEY
in devenv to match backend config key
- Fix timeout middleware to destroy request AFTER response finishes,
preventing truncated 504 responses
- Fix quality=0 parsing to preserve explicit zero (was silently overridden to 85)
- Replace require('fs') with proper ES module import in upload-storage.ts
- Refactor font conversion temp-dir boilerplate into withTempInput helper
- Document FontForge escaping limitations (single quotes only)
- Fix misleading comment in image.ts about sharp metadata decoding
AI-assisted-by: qwen3.7-plus
* 🐛 Fix code review issues in media-processor (round 2)
- Fix queue middleware to skip next() when response already ended,
preventing orphaned work after timeout
- Fix hybrid storage to use disk when Content-Length is absent (chunked
transfer), preventing unbounded memory allocation
- Add source image format validation in generateThumbnail to reject
unsupported formats (TIFF, BMP, etc.) with 400 instead of 500
- Remove dead code in convertFont for unreachable woff→woff path
- Remove unused isEnabled() method from LokiLogTransport
- Fix sfntToWoff to use correct extension (.ttf/.otf) based on source type
- Extract queue middleware to separate file for testability
- Add comprehensive tests for queue middleware and upload storage
AI-assisted-by: qwen3.7-plus
* 🐛 Fix code review issues in media-processor (round 3)
- Fix disk-backed upload cleanup after successful requests by adding
cleanup middleware that removes temp files on response finish/close
- Wrap sharp metadata/decoding errors as 400 validation errors instead
of 500 internal errors
- Only apply flatten() for JPEG output to preserve alpha channel in
PNG and WebP outputs
AI-assisted-by: qwen3.7-plus
* ✨ Add comprehensive tests for media-processor
Phase 1 - Cleanup verification:
- Add cleanup middleware unit tests (6 tests)
- Add HTTP upload cleanup integration tests (5 tests)
Phase 2 - Error handling & alpha preservation:
- Add sharp error wrapping tests (4 tests)
- Add HTTP malformed image tests (2 tests)
- Add alpha preservation tests (3 tests)
Phase 3 - Edge cases:
- Add upload storage edge case tests (3 tests)
- Add queue middleware edge case tests (4 tests)
Phase 4 - Backend mock verification:
- Fix backend mocks to include :mtype field in image info responses
- Verify all error codes match actual service behavior
Total: 27 new tests added (160 tests passing)
AI-assisted-by: qwen3.7-plus
* 🐛 Fix code review issues in media-processor (round 4)
- Add Zod validation constraints for config values (int, positive, min)
- Fix auth middleware to compare Buffer byte lengths instead of string lengths
- Validate requested output dimensions in generateThumbnail (crop mode)
- Change queue middleware to release slot via callback in finally block
- Add comprehensive tests for all fixes
AI-assisted-by: qwen3.7-plus
* 🐛 Close HTTP response streams in backend media remote
- Wrap stream consumption in try/finally with .close() calls
- Add tests to verify stream closure for info, font-convert, and thumbnail
AI-assisted-by: qwen3.7-plus
* 🐛 Fix queue slot leak on upload failures
Make releaseQueue idempotent and attach fallback listener to release
slot when response finishes. This covers Multer errors that bypass
the route handler's finally block, preventing permanent queue stall.
AI-assisted-by: qwen3.7-plus
* 🐛 Cancel processing on timeout
Create AbortController in timeout middleware and abort signal when
timeout fires. Pass signal to Sharp and FontForge to cancel ongoing
processing and release resources when request is cancelled.
AI-assisted-by: qwen3.7-plus
* 🐛 Fix code review issues in media-processor (round 6)
- Error handler: check headersSent before writing response to prevent
ERR_HTTP_HEADERS_SENT when timeout already sent 504
- Timeout config: increase default requestTimeout from 60s to 180s to
match font processing timeout (120s) and backend request timeout
- Image processing: check abort signal before starting Sharp operations
to cancel processing when timeout fires
- Queue lifecycle: remove res.on('close', release) fallback to hold
queue slot until processing completes, preventing concurrency limit
violation when client disconnects
AI-assisted-by: qwen3.7-plus
* 🐛 Close HTTP response stream in download-image
Wrap response body in with-open to ensure stream is closed after
writing to temp file, preventing HTTP connection leaks on repeated
URL imports.
AI-assisted-by: qwen3.7-plus
* 🐛 Close HTTP response stream on validation errors in download-image
Move with-open to wrap the entire validation and processing block,
ensuring the response body stream is closed even when validation fails
(non-2xx status, missing size, invalid media type). This prevents
HTTP connection leaks on repeated failed downloads.
Add test to verify stream closure on validation errors.
AI-assisted-by: qwen3.7-plus
* 🐛 Pass abort signal to Sharp toBuffer for timeout cancellation
Wrap Sharp's toBuffer() with Promise.race to check abort signal during
processing. This ensures large thumbnails stop processing when the
request times out, preventing wasted CPU/memory and queue capacity.
Add test to verify abort during toBuffer operation.
AI-assisted-by: qwen3.7-plus
* 🐛 Hold queue slot until Sharp completes and handle client disconnect
- Remove Promise.race from generateThumbnail — Sharp processing now
completes fully before queue slot is released, preventing concurrency
limit violations under timeout conditions
- Remove res.on("finish", release) fallback from queue middleware —
error handler now explicitly calls releaseQueue in all error paths
- Add res.on("close") handler in timeout middleware to abort signal
when client disconnects, ensuring processing stops early
- Add tests for client disconnect handling and queue slot lifecycle
AI-assisted-by: qwen3.7-plus
* 🐛 Address round 9 review findings
- Document Sharp 0.35.3 cancellation limitation in image.ts
- Add integration test for timeout cleanup with large images
- Fix font tools (sfntToWoff, woffToSfnt, woff2ToSfnt) to throw
ProcessingError on resource limit kills instead of returning null
- Validate font signatures for same-format conversions to prevent
arbitrary files from being persisted as valid fonts
- Fix concurrent mkdtemp race in upload-storage by using shared
initialization promise
AI-assisted-by: qwen3.7-plus
* 🐛 Address round 10 review findings
- Add tmpdir assertion in font.ts to prevent path injection
- Preserve original error in queue middleware catch handler
- Change auth middleware response type from "internal" to "authorization"
- Add cleanup flag to prevent double cleanup in cleanup middleware
- Move quality clamping into parseQuality function for consistency
- Add integration tests for quality parameter clamping at route level
- Update existing tests to match new auth response type
AI-assisted-by: qwen3.7-plus
* 🐛 Address round 11 review findings
- Extract releaseSlot helper in error-handler to reduce duplication
- Remove redundant try/catch in font.ts withTempDir cleanup
- Improve font path validation error message for clarity
- Move path validation before try/catch to prevent swallowing
- Add debug logging for cleanup failures in cleanup middleware
- Inline TransportTargetSpec type alias in logger.ts
- Extract logging middleware to separate file for consistency
- Remove duplicate MIME validation in image thumbnail route
- Add test for font path validation (outside tmpdir rejection)
- Add tests for error handler queue release across all branches
AI-assisted-by: qwen3.7-plus
* 🐛 Remove Content-Length header from multipart requests
The JDK's HttpClient rejects Content-Length as a restricted header,
causing IllegalArgumentException when sending multipart requests to the
media-processor. Remove the explicit Content-Length header and let the
JDK use chunked transfer encoding. The media-processor will use disk
storage for all multipart requests (safe default behavior).
Remove unused size computations (file-size, header-bytes, footer-bytes,
total-size) that were only used for Content-Length.
Update test to verify Content-Length is not present in request headers.
AI-assisted-by: qwen3.7-plus
* 🐛 Fix pino ESM bundling for media-processor
Mark pino and its transports (pino-pretty, pino-loki) as external to
avoid bundling issues with worker thread modules that reference
__dirname (not available in ES modules).
AI-assisted-by: qwen3.7-plus
* ♻️ Rename nitrate config to admin-console
Rename user-facing configuration from 'nitrate' to 'admin-console':
- Feature flags: :nitrate -> :admin-console, :nitrate-bulk-create-profiles -> :admin-console-bulk-create-profiles
- Config keys: :nitrate-shared-key -> :admin-console-shared-key, :nitrate-backend-uri -> :admin-console-uri
- Shared-keys map entry: :nitrate -> :admin-console (setup.clj + main.clj)
- Env vars: PENPOT_NITRATE_SHARED_KEY -> PENPOT_ADMIN_CONSOLE_SHARED_KEY, PENPOT_NITRATE_BACKEND_URI removed (consolidated into PENPOT_ADMIN_CONSOLE_URI)
- Docker/nginx: PENPOT_NITRATE_URI -> PENPOT_ADMIN_CONSOLE_URI
Code namespaces, file paths, CSS classes, and i18n keys stay as-is.
AI-assisted-by: mimo-v2.5-pro
* ♻️ Rename initialize-user-in-nitrate-organization to initialize-user-in-organization
Part of the nitrate -> admin-console rename series. The function and all 9 references across 6 files have been renamed.
* ♻️ Rename :nitrate-bulk-create-profiles-not-allowed to :bulk-create-profiles-not-allowed
* ♻️ Inline nitrate-permissions into app.common.types.organization
- Delete app.common.types.nitrate-permissions and its test
- Move permission rules (allowed?, can-send-invitations?, etc.) into organization.cljc
- Harmonize all consumers to use alias cto for app.common.types.organization
- Update test runner and create organization_test.cljc
Remove the guard in stop-devenv that refused to stop ws0 while any
ws1+ instance was running. Each workspace is now fully independent
and can be started/stopped in any order. Shared infra shuts down
only when no instances remain running.
Updated docs (devenv.md, agentic-devenv.md) and devenv memory to
reflect the new behavior.
AI-assisted-by: mimo-v2.5-pro
Move Dockerfile.frontend, .backend, .exporter, .mcp and .storybook
under docker/images/ from ubuntu:26.04 / nginx-unprivileged / a
manual Node tarball install to Docker Hardened Images (Debian 13,
or Alpine for storybook). storybook and mcp get a true non-dev
runtime; frontend, backend and exporter keep the -dev tag as their
final image, since each needs a shell and/or package manager at
container runtime (nginx templating, fontforge/python3, and a
headless-browser stack, respectively).
