Supersedes #1353, per the discussion in #1352: one `API_KEY` environment
variable; when set, a request carrying it is authenticated as the first
admin, and cookie sessions keep working exactly as before (cookie **or**
key). Nothing else changes. No table, no UI, no settings-tab switch, no
per-user keys.
## What
- `API_KEY` (env). Unset → the feature is off and none of the new code
runs.
- `Authorization: Bearer <key>` or `X-Api-Key: <key>` on any existing
`/api/*` route authenticates that request as the first admin in
`users.db` (`ORDER BY id`), or as a bare admin identity
(`user_id="api"`, `is_admin=True`, no local user row) if the install has
no admin yet. Per request only; nothing is persisted; the admin's role
is read live, so deleting or demoting that user takes effect on the next
request.
- Both headers are checked and either may match. That is what makes the
key usable behind a reverse proxy that injects its own `Authorization`
header (oauth2-proxy, Authelia, forwardAuth): send the key in
`X-Api-Key`.
- A credential that is **not** the key is ignored and the request
continues on the normal session path, so proxy-forwarded tokens are
unaffected. Without a valid session such a request gets the usual `401
{"error": "Unauthorized"}`, identical to a request with no credential,
so there is nothing to probe.
## How
- `shelfmark/config/env.py`: `API_KEY = os.getenv("API_KEY",
"").strip()`.
- `shelfmark/core/api_key.py`: `extract_api_key_candidates()` (Bearer
token if the scheme is Bearer, then `X-Api-Key`) and `matches_api_key()`
using `hmac.compare_digest` on bytes.
- `shelfmark/core/user_db.py`: `UserDB.get_first_admin()`.
- `shelfmark/main.py`: `api_key_auth_middleware` (`before_request`,
registered before `proxy_auth_middleware`, which early-returns for keyed
requests). Only `/api/` paths; `/api/health` and `/api/auth/*` exempt;
no-op when `API_KEY` is unset or the auth mode is `none`. On a match it
mirrors the proxy-auth pattern: `session.clear()` then populate
`user_id` / `is_admin` / `db_user_id` for this request, `permanent =
False`, `modified = False`, `g.api_key_auth = True`. An `after_request`
hook guarantees no `Set-Cookie` is written for a keyed request even if a
handler dirties the session.
- `docs/api-access.md` (new), the `API_KEY` entry in
`docs/environment-variables.md`, and a README link.
## Security
- Constant-time compare; the key is never logged or echoed.
- Keyed requests never mint or refresh a session cookie and ignore any
cookie sent with them (a non-admin cookie plus the key yields admin for
that request; the browser's own session is left untouched and usable).
- The mismatch path touches neither the session nor `g`, so a stray
bearer on a browser request can neither log the user out nor change how
their cookie is refreshed.
- Store errors during the admin lookup fail closed (`500 {"error":
"Authentication error"}`), never to anonymous.
- Verified against Flask's `save_session` / `should_set_cookie`
ordering, and under auth modes `none`, `builtin`, `proxy`.
## Tests
`tests/core/test_api_key_env.py` (36): extraction and matching;
first-admin lookup; middleware behaviour on a guarded route and an admin
route, with and without a user_db, `X-Api-Key`, both-headers
combinations, no `Set-Cookie` when a handler dirties the session,
incoming non-admin cookie ignored, browser cookie still usable after a
keyed request, security headers, store error → 500, mismatch → guard's
401 / cookie path / permanent cookie untouched, unset → off, exempt
paths and path probes, `none` and `proxy` modes, deleted and demoted
first admin, a keyed write passing the guard. Existing auth suites
unchanged. All CI gates green on the fork:
https://github.com/gavinmcfall/shelfmark/pull/2 (CI-only draft).
Also exercised against a running instance: 47 scripted checks including
150 concurrent requests, proxy-mode switching through the key, an
unset-key restart, and a log scan for the key.
## Naming
`API_KEY` as discussed. If you'd rather namespace it
(`SHELFMARK_API_KEY`) to avoid clashing with other tools' env vars in
shared compose files, it is a one-line change; say the word.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Hardcover's August 2026 token system replaced blanket access with
per-token scopes, and nothing in the docs said which ones Shelfmark
actually uses. A key missing write:library or write:lists still passes
Test Connection -- the reading-status and auto-remove-on-download calls
just fail silently afterwards.
Verified against a live hc_pat_ key: every scope in the table backs a
query or mutation the provider really issues, and the omitted ones
(journal, goals, reviews, prompts, notifications, account) are absent
from the provider entirely.
Refs #1240
The orphan sweep ran a container-wide 'pkill -9 -f
chrome|chromium|Xvfb|ffmpeg', so it also matched browsers another bypass
was still driving. Scope it by process group: kill only our own group
and groups whose leader has died. Spawn the helper with
start_new_session so its browser tree is identifiable, tear the whole
group down after every run (a timed-out helper used to leak its Chrome
and Xvfb), and have an orphaned helper take its browser down with it.
Fixes#1231.
LOG_LEVEL never reached the app logger: env.py hardcoded the level to
DEBUG or INFO, so INFO lines kept appearing under LOG_LEVEL=error. Read
it from the env var and advanced settings, normalize unknown values to
INFO, and expose it as a setting. entrypoint.sh now normalizes
gunicorn's level too, so a typo falls back to info instead of stopping
the container from booting.
Z-Library gates the first hit on /md5/<hash> with a 503 whose only
payload is a Set-Cookie; echoing that cookie back returns the 302 to the
real page. html_get_page dropped it and re-ran the same rejected request
on every retry, ending in "No download URL resolved". Retry once with
the
cookies the 503 issued.
Fixes#1185Fixes#1187
Hardcover forwards the `sort` argument to Typesense's `sort_by` and
rejects
the entire search if it dislikes the value -- an unknown field, a bare
field
name with no direction, or more than three sort keys. A rejected search
is
not a GraphQL error: it comes back as HTTP 200, no `errors` key, and a
null
`results` body.
_extract_typesense_hits() reads that null as `hits=[], found=0`, so a
failed
search was indistinguishable from one that matched nothing. Users saw
zero
results with a healthy container and no log line explaining why.
Add _execute_search_query(), used by the three sort-bearing call sites
(book
search, field typeahead, series resolution):
- Detect the rejection via the null `results` body. A search that
genuinely
matched nothing still returns a results object with `found: 0`, so empty
result sets are not mistaken for failures.
- Retry once with an empty sort, which Hardcover always accepts, so
searches
return results instead of nothing.
- Keep that fallback sticky for 15 minutes so every subsequent search
does
not pay for a request known to fail, and let it expire so sort order
comes
back on its own if the index is fixed upstream.
- Log rejections that no sort can explain, and retries that also fail,
at
ERROR instead of discarding them.
While the fallback is active, results fall back to Typesense's default
ordering regardless of the selected sort. Degraded ordering beats no
results,
and it is now logged rather than silent.
SORT_MAPPING itself is unchanged: all five of its values were verified
against
the live API and return results. The `sort: "relevance"` reported in
#1179 was
the raw SortOrder value sent by v1.3.5; the mapping already fixed that.
What
remained unfixed, and is fixed here, is that the failure was invisible.
Fixes#1179
- Updated mirror selection
- Removed built-in mirror options, users must provide their own
configurations
- Set Universal search to default, added ability to disable direct
source
- Updated documentation
- Updated makefile
- Adds a non-root startup path at user 1000:1000 - skips privilege
escalation and ownership checks. Works e.g. for kubernetes deployments
(user 1000:1000 and runAsNonRoot enabled).
- Remove startup check/chown commands for user-owned folders. Checks can
be done with a "Test destination" button in settings which performs a
test write. Users are responsible for fixing their own permissions.
- Update docs
## Summary
Adds AudiobookBay as a web-scraping release source for audiobook
torrents. Once enabled, a new tab shows up in the Find Releases modal.
## What's New
- **AudiobookBay source** – Search AudiobookBay for audiobook torrents
from the Shelfmark UI
- **Torrent downloads** – Extract magnet links from detail pages and add
them to the configured torrent client
- **Audiobook-only** – Source is limited to audiobooks
- **Download clients:** Currently uses the torrent client configured
under **Prowlarr > Download Clients**.
- Audiobook-specific categories (e.g. `QBITTORRENT_CATEGORY_AUDIOBOOK`)
are applied when set.
- **Settings → AudiobookBay**:
- Enable toggle
- Hostname
- Max pages to search (default 5)
- Rate limit delay in seconds (default 1)
## How It Works
1. User searches for an audiobook; AudiobookBay is queried if enabled.
2. Results show title, language, format, and size
3. User selects a release; the handler fetches the detail page and
extracts the magnet link.
4. Magnet link is sent to the configured torrent client.
## Testing
- Unit tests for source, handler, scraper, and utils
- Mocked HTTP requests and torrent client calls
- Coverage for search, relevance filtering, language mapping, size
parsing, and download flow
### Screenshots
<img width="600" alt="image"
src="https://github.com/user-attachments/assets/2e10a259-5c35-4065-980d-b59a1c961c9f"
/>
- Adds a comprehensive multi-user request system to the existing
download flow
- Request configuration is policy based. Configure global settings for
content type, or narrow down policy for specific sources (E.g. allow
direct downloads, set prowlarr to request only, block IRC completely,
etc).
- Global policy configuration and per-user overrides for tailored
configs
- Replaced downloads sidebar with ActivitySidebar, combining active
downloads with requests. Admin management of user requests is done here,
and admins have view of downloads from all users. Sidebar can now be
pinned.
- Request either a standard book or a specific release. Release-requests
are used if you permit one source differently than the other. On
book-level requests, admins pick the specific file to be attached to the
fulfilled request.
- Users can request books with a note
This is WIP so some features are still not complete (notifications, more
automatic release selection, among others).
Closes#552
## Summary
Adds OIDC authentication and multi-user support to Shelfmark. Users can
now be managed individually with per-user download settings, while
maintaining full backwards compatibility with existing auth modes
(no-auth, builtin, proxy, CWA).
### Authentication
- **OIDC login** with PKCE, auto-discovery, group-based admin mapping
- **Password fallback** when OIDC is enabled (prevents admin lockout)
- **Auto-provisioning** of OIDC users (configurable on/off)
- **Email-based linking** of pre-created users to OIDC accounts
- **Lockout prevention** — requires a local admin before OIDC can be
enabled
### User Management
- **SQLite user database** (`users.db`) with admin CRUD API
- **Users management tab** in settings UI (admin-only)
- **Settings restricted to admins** in multi-user modes (builtin/OIDC) —
non-admin users cannot access settings
- Create, edit, and delete users with role assignment (admin/user)
- Password management for builtin auth users
- OIDC users shown with provider badge (password fields hidden)
- Per-user configurable settings:
- **Download destination** — custom folder path per user
- **BookLore library & path** — dropdown select, each user's books go to
their own library
- **Email recipients** — per-user email delivery targets
- **`{User}` template variable** — use in destination paths (e.g.,
`/books/{User}/`)
- Settings override model: per-user values override globals, empty/unset
falls back to global defaults
### Download Scoping
- **Per-user download visibility** — non-admins only see their own
downloads
- **Username display** in downloads sidebar (shows who requested each
download)
- **WebSocket room-based filtering** — admins see all, users see only
their own
- **Download progress scoping** — progress events routed to correct user
rooms
### BookLore Integration
- **Dynamic dropdown selects** for library/path (replaces text inputs)
- **Per-user library/path overrides** via user settings
- **Options cache refresh** after Test Connection
### Security
- SQL injection prevention (column whitelist on user updates)
- Generic OIDC error messages (no internal detail leakage)
- Admin self-deletion and last-local-admin deletion guards
- OIDC role overwrite fix (only updates role when admin_group is
configured)
## Migration
**No migration script needed.** The `users.db` is created automatically
on first startup. Existing builtin auth users are auto-migrated to the
database on their first login. All other auth modes (no-auth, proxy,
CWA) continue working unchanged.
## Test Plan
- [x] All 519 tests passing, 0 failures
- [ ] Test no-auth mode: settings accessible, downloads work without
login
- [ ] Test builtin auth: legacy credentials auto-migrate on login, new
users can be created
- [ ] Test OIDC auth: login flow, callback, auto-provisioning,
group-based admin
- [ ] Test CWA auth: unchanged behavior
- [ ] Test proxy auth: unchanged behavior
- [ ] Test per-user downloads: non-admin sees only own downloads
- [ ] Test BookLore dropdowns: library/path selection, per-user
overrides
- [ ] Test Docker build: no Dockerfile changes needed
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
- Base URL option in settings for reverse proxy setups
- Fix NZB downloads not deleting on completion
- Fix handling for audiobook files over 100+ parts
- Fix prowlarr search timeout
- Fix prowlarr categorisation for expanded searches
- Changes the auth settings to support more than two auth types
- Added a proxy auth type with settings for user and optionally group
headers
- Added a global middleware `proxy_auth_middleware` to handle proxy auth
(it does nothing if any other auth mode is set)
- Added support for proxy auth to `get_auth_mode`, `login_required`,
`api_login/out`, and `api_auth_check`
- Added a backend check to make protect the API for settings when admin
is required
---------
Co-authored-by: Joshua Tag Howard <git@jthoward.dev>
Co-authored-by: Alex <alex.bilbie1@gmail.com>
- Added alternative file processing mode. Save files directly into a
library folder and set up file names / directories based on user
preference.
- Uses template based naming and directory creation. E.g. {Author} /
{Series} {Title} {Part} etc. Works for saving correctly to libraries
such as Audiobookshelf.
- Use torrent hardlinking directly into library directories.
Hey, made the tweaks we discussed, plus a couple related fixes :)
- Removed APP_ENV entirely. All dev-specific functionality is enabled
via `DEBUG: true` env var
- Set secure cookie handling to false by default, added to the readme to
enable if exclusively using HTTPS connection
- Fixed healthcheck potentially not working with auth enabled
- Removed APP_ENV from docker compose files and made sure app.db lines
are included in all versions.
APP_ENV in people's existing composes should get ignored entirely and
will be put on the default env, so no issues when updating.
Re: Issue #122
Fetches content type from search results - displays it on thumbnails in
results grid;
Fetches content type from book id detail page (dfaults to "Other") and
uses it to construct the `final_path`.
---------
Co-authored-by: Patricia Ritter <pritter@events.com>
Co-authored-by: CaliBrain <calibrain@l4n.xyz>
This commit removes all references to the deprecated /request route
prefix
that was previously used for dual routing. The following changes were
made:
- Removed register_dual_routes() function that registered routes with
/request prefix
- Removed url_for_with_request() helper function for generating /request
URLs
- Removed call to register_dual_routes(app) at application startup
- Removed /request/ prefixed favicon routes
- Updated StatusEndpointFilter to remove /request/api/status log
filtering
- Removed unused flask_url_for import
All routes now only use the standard paths without the /request prefix.
This PR was coauthored by alexhb1 and davidemarcoli. It builds on the FE
rework created by alex, but adds a myriad of additional tweaks and
optimizations to make the frontend feel modern, fast, and responsive.
The summary of the changes is as follows:
### Architecture Changes
React/TypeScript Migration: Refactored frontend from template/JS
structure to React/TypeScript application for better maintainability and
scalability
WebSocket Integration: Implemented real-time updates for download status
and progress with automatic fallback to polling
Gevent Worker: Configured production WebSocket support
### UI/UX Improvements
<img width="1502" height="890" alt="Screenshot 2025-11-10 at 10 02
59 AM"
src="https://github.com/user-attachments/assets/86bf8649-623f-413c-b8e5-656e687e55a8"
/>
Downloads Sidebar: Replaced bottom downloads section with sidebar
interface for better organization
<img width="201" height="450" alt="Screenshot 2025-11-10 at 10 07 52 AM"
src="https://github.com/user-attachments/assets/92b98e7c-c3bc-4b7e-80f1-252c3a760e33"
/>
Status Badges: Color-coded download status indicators instead of plain
text
Pinned Header: Fixed header position for consistent navigation
Enhanced Book Cards: Improved layout and hover states with info modal
button
<img width="1474" height="899" alt="Screenshot 2025-11-10 at 10 08
18 AM"
src="https://github.com/user-attachments/assets/9216d8a3-f662-434d-80e6-2a69b96abc31"
/>
Download Progress: Circular progress indicator on download buttons
Toast Notifications: Added user feedback for actions
Spinner Feedback: Loading indicators on search and download buttons
Animations: Smooth transitions and fluid progress updates
### Mobile & Responsive Design
Mobile-friendly Layouts: Optimized book cards and search interface for
mobile
<img width="225" height="450" alt="Screenshot 2025-11-10 at 10 05 49 AM"
src="https://github.com/user-attachments/assets/c8236c1c-5837-4309-9577-46db7292a54b"
/>
Keyboard Handling: Improved mobile keyboard behavior with proper input
types
PWA Improvements: Enhanced progressive web app functionality
Responsive Search: Better search box width and positioning across
devices
### Developer Experience
Development Mode: Separate frontend dev server that works with existing
backend container
Makefile: Added build automation and development commands
Documentation: Updated README with frontend architecture details
### Bug Fixes
Fixed "Clear completed" functionality
Fixed dark mode toggle text
Fixed sticky header behavior
Fixed mobile search box positioning
Removed active downloads requirement for initial state view
### Additional Features
ESC Key: Close downloads sidebar with ESC key
Calibre-Web Button: Direct link to Calibre-Web instance
<img width="282" height="83" alt="Screenshot 2025-11-11 at 9 38 05 AM"
src="https://github.com/user-attachments/assets/273075be-9743-4e13-9e48-5bf498f6c067"
/>
Granular Status Tracking: More detailed download progress information
obtained via websockets
---------
Co-authored-by: Alex <alex.bilbie1@gmail.com>
Co-authored-by: Zack Yancey <yanceyz@proton.me>
Co-authored-by: davidemarcoli <davide@marcoli.ch>
Adding support for an external CloudFlare bypasser service and
introducing a new Docker image build with a dedicated target.
Key Changes
- Added `cloudflare_bypasser_external.py` for external bypasser
integration.
- Updated Docker Compose files to support the new service.
- Introduced a new Docker target for building a separate image for the
external bypasser.
- Refactored relevant modules to utilize the external bypasser when
configured.
- Documentation and configuration updates to reflect new options and
Docker targets.
Impact
- Users can now choose between internal and external CloudFlare
bypassing.
- New Docker image and target streamline deployment of the external
bypasser.
- Improved modularity and maintainability.
- No breaking changes for existing workflows.
Testing
- Manual and E2E tests performed for both bypasser modes.
- Docker Compose setups and new image build verified for development and
production.
Notes
Please review the new configuration options and Docker targets. Update
your environment and deployment scripts as needed. Feedback and
suggestions are welcome!
Added details to fix "database locked" error for anyone using a cifs
share for their library under volumes. nobrl option needs to be added to
the mount line in your fstab file. e.g. //192.168.1.1/Books /media/books
cifs credentials=.smbcredentials,uid=1000,gid=1000,iocharset=utf8,nobrl
This commit adds user authentication, using Calibre-Web's app.db as its
authentication source, as requested in #56. It uses @prinzpi's
[comment](https://github.com/calibrain/calibre-web-automated-book-downloader/issues/56#issuecomment-2919335169)
as a starting point, but integrates the logic directly into the app.
This requires the user to specify the environment variable CONFIG_ROOT,
set to Calibre-Web's config directory (the directory that contains the
app.db database that holds the user's authentication information).
If the user does not wish to add authentication, they can simply leave
CONFIG_ROOT unset, or not pointing at Calibre-Web's app.db directory.
Refactor: Improve Docker build, add Tor support, use SeleniumBase
- Overhauled Dockerfile:
- Switched to python:3.10-slim base.
- Implemented multi-stage builds (base, standard, tor).
- Consolidated RUN layers for efficiency.
- Added locale/timezone setup.
- Added Tor setup and iptables configuration in dedicated stage/script.
- Replaced DrissionPage with SeleniumBase for Cloudflare bypassing.
- Added Tor support via `docker-compose.tor.yml` and `tor.sh` script.
- Updated `docker-compose.yml` to build locally and changed default port
to 8083.
- Added `docker-compose.dev.yml` and `docker-compose.tor.dev.yml`.
- Updated `entrypoint.sh` for timezone and sudo usage.
- Added/Updated environment variables (`TZ`, `USING_TOR`, etc.).
- Improved `.dockerignore`.
- Updated `readme.md` to reflect port changes, build process, document
`TZ`, and add details about the new Tor variant.
Gavin McFallandClaude Fable 5.1
CaliBrain
Alex SchittkoandCaliBrain
Alex
bonsai-dreams
Michael Joshua SaulandClaude Opus 4.6
Webysther Sperandio
Alex
Federico Della Rovere
smoke007
Timothy Allen