claude.ai's connector dialog does offer request headers, on its second step,
after the URL has been probed, so the connector no longer needs the secret
path. MCP_AUTH_TOKEN already worked there as a bearer or X-Api-Key, but it
also opens /api, which can replace the Schulcloud token and stream the file
mirror, and claude.ai stores the header's value.
MCP_CONNECTOR_TOKEN is a second token, accepted on /mcp only and refused on
/api, and rotated without touching Claude Code or the CLI. The config refuses
one shorter than 32 characters, equal to MCP_AUTH_TOKEN, or set without it,
and never echoes a value. Every accepted token is compared in full, so the
timing does not tell which one matched.
The gate also takes a bare Authorization value, because claude.ai sends a
header exactly as typed and its docs warn that most servers reject a token
entered without "Bearer ". It takes X-Auth-Token too, the other name its
dialog offers.
The docs now set up the header; the secret path stays as a fallback for
clients that cannot send one. 184 tests. Smoke 79/79 and 77/77 on the local
instance.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
claude.ai's connector dialog takes a name and a URL. Sending a bearer token
needs a "Request headers" beta most accounts lack, and OAuth is not built
yet, so with MCP_PATH_SECRET set the endpoint is also served at
/<secret>/mcp without the bearer token — a trial until OAuth replaces it.
The path is the credential there. It is compared in constant time, and a
wrong one answers 404 like any unknown path. The config refuses fewer than 32
URL-safe characters and never echoes the value, nothing in the server logs
request paths, and the Caddy snippet rewrites the segment before an access
log entry is written (verified against Caddy 2.11). Claude Code and the CLI
keep the bearer token; DEPLOYMENT.md says what the path trades away.
178 tests. Smoke 76/76 and 74/74 on the local instance.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A token lasts 30 days and only a browser login yields one — the account is
federated, so the server cannot mint it. Replacing it meant editing .env and
recreating the container, every month.
`schulcloud token set` (a hidden prompt, or piped input) and a /token page
both send it to PUT /api/token. The server checks it with Schulcloud first —
well-formed, unexpired, still logged in, the same account — then swaps it
into the config every request reads, restarts the keepalive and saves it in
STATE_DIR, a new volume, with mode 0600. At startup the newer of the saved
token and TSC_JWT_COOKIE wins, unless they belong to different accounts. A
refused paste changes nothing, and the token is never logged.
The keepalive's pings carry a generation, so a 401 for the old token that
arrives after a swap cannot stop the new cycle. `schulcloud token`, whoami
and the log report the expiry and warn a week ahead.
Found on the way: a host that is off for more than two hours loses the
session however long the token has left — this machine lost it overnight —
which is what the always-on Pi is for.
174 tests. Smoke 72/72 on the local instance, and a real swap verified end to
end there.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A course or room can now be attached to a message rather than fetched:
schulcloud://courses/<id> and schulcloud://rooms/<id> carry exactly what
get_course and get_room return. Deliberately coarse — a picker lists every
resource at once, which suits some twenty courses and not a thousand files.
Two prompts, in German because the school is: zusammenfassung summarises a
course or room, and pruefungsvorbereitung prepares for an exam with practice
questions and a study plan. Each embeds the overview and says where material
hides and what cannot be read.
Claude Code shaped the details, read from its bundle rather than its docs.
It splits prompt arguments on whitespace and drops extra words, so words
arrive joined with "_", and courses match by fragments, whole words first,
so LF1 is not ambiguous with LF10. Its @ autocomplete shows a resource's
description, so the description carries the name. Errors are ProtocolError,
because McpError's message prefix is doubled by the client.
Verified in interactive Claude Code: @-mention, autocomplete and the prompt
commands. 157 tests. Smoke 67/67 live; 69/69 and 67/67 on the local instance.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Many teachers never use topics or boards; their material sits in the
course's file area, and the tools answered "0 files" for courses holding
dozens of worksheets — 21 of 26 courses on the live account. Persönliche,
Kurs-, Team- and Geteilte Dateien live in the legacy file store, not in
files-storage, and its service is not in the public ingress. The only way in
is the legacy client: HTML listings, and GET /files/signedurl for a
pre-signed download.
core/legacy-files.ts turns that into one path tree — /my, /courses/<course>,
/teams/<team>, /shared — resolving names that contain "/", ids anywhere in a
path, and wrong or ambiguous names with a message saying what is there. A
listing that does not parse throws; it never reads as an empty folder.
Some of the legacy client's GET routes write (GET /files/share/ mints a
share token), so getFileManagerPage allows only the listing routes, by
pattern. Signed URLs are fetched with no credentials and must be https.
- MCP: fs_list, fs_tree, fs_find and fs_read; get_course lists course files.
- CLI: schulcloud fs ls, tree, find and get, recursive and resumable.
- API: /api/fs/list, tree, find and file.
- Index: the crawl walks the file manager (INDEX_FILE_MANAGER, on by
default), so search covers the text inside those files and sync mirrors
them under <course>/Kurs-Dateien.
The local instance gains a fixture for all four areas. It needed a loopback,
so signed URLs open from the host, and a pre-created bucket, since MinIO
does not implement PutBucketCors.
135 tests. Smoke 55/55 live; 57/57 and 55/55 on the local instance.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
With .env pointing at the real account, testing against the local instance
became dangerous: process env beats --env-file, so a local smoke run would
crawl fixtures straight into the live index, where a per-course refresh
carries them forward indefinitely. mcp-env.sh now pins its own database
(schulcloud_local) and mirror as well as the instance.
The override publishes the server on MCP_HOST_PORT and takes
CRAWL_INTERVAL_MS from .env, since what_changed can only report what
happened between crawls. The build context leaves out tmp/, which holds a
mirror of the account's files, and local-instance/.
probe checks live what the gap fixes established locally: the /api/v1 course
and user routes, classes, room allowedOperations as an object, and the
preview enums on a real file. A failure there means a feature degrades
rather than breaks.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Every area — courses, rooms, boards, topics, tasks, files, quizzes, teams,
groups, submissions, grades — was checked for data the instance has and the
tools did not show.
Grades and feedback. A teacher's /homework page is a different page from a
student's: grade and comment live in the grading form, one block per
submission, so a teacher account reported every graded submission as having
neither. parseTeacherGrading reads the form, and list_submissions can now
include the written feedback and who handed the work in.
Names. /api/v1 is partly served: courses, users and classes survive in the
deployment's ingress table, and users/{id} is the only route from an id to a
name. Submitters, file creators and course teachers resolve through it, and
degrade to "not visible to this account" where a student may not read them.
Courses, rooms and classes. get_course adds the description, teachers,
member count and weekly timetable from /api/v1/courses. list_classes is new.
get_room reports what the account may do — allowedOperations is an object of
booleans, not the list it was typed as — and applicants and invitation links
where it may manage them.
Board and topic content. Link descriptions, image alt text, drawing and
video-conference titles, the ids behind external tools and H5P content (the
only thing resembling a quiz), and what a deleted element used to be. Topic
Etherpad pads are read like board pads, and htmlToText keeps table columns
apart and drops template indentation.
Files. A scan with no text layer falls back to the preview endpoint, whose
width and outputFormat are undocumented enums, so Claude gets a picture of
the page; list_files reports counts and sizes. Teams stay documented as
unreadable at any API version; their files come later.
What the crawl missed. Tasks attached to topics (18 of 60 on the live
account), each course's own file area, and — behind INDEX_PERSONAL_FILES —
personal files and submissions with their grade comments, so search and
what_changed cover grading. A submission hit points at get_task.
The local instance's preview profile gets an ImageMagick policy that allows
the coders its 7.1.2 build needs; the image's own denies them all.
110 tests.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A topic page makes the legacy client set several large cookies at once —
jwt plus the Etherpad session it asks for — which overflows nginx's default
4k header buffer. The upstream answered 200 and the proxy still returned 502
with "upstream sent too big header".
It went unnoticed because the MCP server's topic-page scrape degrades to an
empty list by design: the task ids it recovers from that page silently
vanished instead of failing. Changed in the generator and in the generated
config.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The account this was built against is in no rooms, so the whole space was
invisible and easy to dismiss as an empty endpoint. It is not empty in
general — the user had rooms until a teacher removed access — and the
UI's naming actively hides the distinction: the sidebar's *Kurse* entry
links to `/rooms/courses-overview` and lists courses, while *Räume* links
to `/rooms` and lists rooms. A url containing `/rooms` identifies neither.
list_rooms and get_room cover the latter. A room holds boards and nothing
else, so get_room lists boards for get_board (which already reports "in
room" from the board context) plus who else is in it. Room boards report
`isVisible`, which the course-page projection does not, so a draft is
named as a draft instead of being offered and then answering 403.
Rooms also go through the crawl, or they would have become the next
blind spot: their boards are indexed, searchable by both the index and
the live-crawl path, diffed by what_changed, and mirrored by the CLI
under the room's name. The board traversal and the snapshot matcher are
now shared between courses and rooms rather than duplicated, which also
fixed the live-crawl path silently not searching pad contents.
The CLI needed no new command — it is file-centric and inherits rooms
through the manifest — but `--course` now accepts a room id, and says so.
`kind` gains 'room'; the column is plain TEXT, so no migration. 112 tests.
Smoke: 42/42 and 44/44 local, 41/41 and 43/43 live.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
`scripts/simulate-teacher.mjs` creates, edits and deletes what teachers
create — course, room, topic, task, board, columns, cards, rich text,
link, Etherpad pad, folder, files — so the MCP server can be exercised
against content the real account has never held. It is the only thing
here that writes to a Schulcloud and refuses any non-localhost address.
`scripts/mcp-env.sh` points the server and CLI at the instance.
Two gaps it exposed in the stack itself:
The files-storage AMQP consumer is a separate entrypoint, and we were
running only the HTTP one. Nothing was bound to the `files-storage`
exchange, so `TaskService.delete` — which awaits deleteFilesOfParent
over AMQP before touching the task — hung until the request timeout.
Deleting any task or topic answered 408 with the entity still there.
The demo data is dated 2017-2018 and the v3 endpoints filter on those
dates, so a student saw no tasks at all. seed.sh now brings courses and
homework into the present, which is the difference between a fixture
that exercises the student-facing surface and one that looks empty.
Teams turn out to be uncreatable through the API (the legacy service
registers no `create`, v3 has no route), and a new board is unpublished
and 403s for students, so the simulation adopts a seeded team and leaves
one board a draft on purpose.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The legacy client requests an Etherpad session on every topic page whose
lesson has contents, without ever checking whether the lesson contains a
pad — controllers/topics.js collects `etherpadPads` and then ignores it.
With Etherpad unreachable the request fails, `validUntil` comes back
undefined, and `new Date(undefined * 1000)` makes Express reject the
session cookie: "option expires is invalid", rendered as a 500.
So Etherpad was only nominally optional. Moving it out of the `tools`
profile also matches the live deployment, which always runs it.
The topic pages that still 500 are courses the signed-in user is not a
member of; the same page returns 200 for its own teacher. That is the
legacy client rendering a 403 as a 500, upstream behaviour we don't own.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A Docker Compose stack that runs the deployed images
(quay.io/schulcloudverbund/*, thr theme, tag 33.40 — the versions
schulcloud-thueringen.de reports) rather than a rebuild of main, so what we
develop against is the deployed artefact. It exists to produce the states we
can otherwise only observe read-only: log in as the teacher, grade, then read
it back the way the MCP server does.
Faithful where it matters and honest where it isn't:
- Feature flags in env/api.env are a replay of GET /api/v3/config/public from
the live instance, not a hand-picked set; instance identity mirrors the thr
group_vars from dof_app_deploy.
- The proxy is generated from the deployment's own ingress table
(scripts/gen-proxy-conf.py) so the legacy-client / SPA / API path split
matches production; getting it wrong tests a different application.
- Valkey runs in `single` mode so the JWT whitelist expires sessions the way
production does, rather than the in-memory shortcut that hides it.
- No external OAuth / Schulportal login (excluded by request and not
reproducible locally), no BigBlueButton; each divergence is marked at the
line it affects. Everything binds to 127.0.0.1 and uses the upstream dev
credentials, which are public.
Profiles keep the heavy pieces opt-in: `tools` adds Etherpad/H5P/tldraw/
Collabora, `av` adds ClamAV, `preview` adds thumbnailing.
seed.sh loads the upstream demo school (the same call the deployment's init job
makes) and registers MinIO as the legacy storage provider, which has no seed
data on purpose. The demo data already contains the grading states that are
hard to obtain from the real account — a feedback-only grade and a 100% one —
which is what surfaced the past-due submitted-text scrape gap.
One config finding baked in: file-storage and h5p validate a token's
issuer/audience against JWT_DOMAIN (default "localhost"), while the API stamps
SC_DOMAIN; without keeping them equal, the homework page's file lookups 401.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
