Last updated: 17 October 2026
Setsu (formerly ClipLib) is developed and operated by an individual developer based in Germany:
Fabian Silbermann Germany Email: fabiansilbermann1@gmail.com
For any privacy question or to exercise your rights (see section 10), email the address above.
Most of what Setsu does happens entirely locally and never reaches us or any third party:
- Your video clips and any exported/trimmed copies.
- Per-clip metadata sidecars (stored next to your clips in a
.clip_metadatafolder): custom name, trim points, playback speed, volume, track state, tags, and a.gameinfofile containing the recorded window/game title and, if applicable, the Discord voice roster (see section 5). - App settings (
settings.json): your clip folder location, feature toggles, and keybindings/controller mappings. - A local activity log (
activity_logs/…): a record of actions you take in the app (renames, trims, speed/volume changes, tag edits, and when you shared a clip). This log stays on your machine and is never uploaded automatically. It may contain clip names and your sharing history. - Minor UI state (e.g. whether the sidebar is pinned).
You can delete any of this at any time by deleting the corresponding files or your clips.
Setsu includes an optional sharing service. Nothing is uploaded until you sign in and choose to share.
Signing in. Accounts use Discord login. When you authenticate, the sharing service
receives and stores your basic Discord identity: display name, username, Discord user ID,
and avatar. Your access token is stored encrypted on your device using your operating
system's secure storage (Windows DPAPI / OS keychain via Electron safeStorage).
When you share a clip, we upload and store:
- the video file itself (trimmed export, up to 500 MB per clip);
- the title and tags you gave it;
- the Discord user IDs of anyone you @mention on the clip.
Profile & social features. You may upload a profile banner image, and the service supports a social feed (listing clips, reactions, comments) and a directory used for @mention autocomplete (which exposes users' id, username, display name, avatar, and Discord ID to other signed-in users).
Legal basis (GDPR Art. 6): your consent and the performance of the service you requested, Art. 6(1)(a) and 6(1)(b). You can stop at any time by not sharing, or by deleting your shared clips/account (contact us if you need help removing content).
To keep the app current, Setsu checks GitHub
(api.github.com / github.com/yuma-dev/setsu) for the latest release and, if you
choose to update, downloads the installer from there. These requests send only a generic
User-Agent header, no account, no identifiers. GitHub, as the host, may log the
request under GitHub's Privacy Statement.
Setsu's recorder (Setsu Recorder, formerly clipdip) can, if you connect it to
Discord, read who is in your Discord voice channel at the moment a clip is recorded,
participants' Discord IDs, usernames, display names, and avatars, and store that roster
locally in the clip's .gameinfo sidecar. This powers @mention autocomplete and the
sharing "mentions" feature.
Important points:
- This is third-party personal data (other people in your call). It stays on your device and only leaves it if you share a clip and @mention those people (see section 3).
- Connecting to Discord uses Discord's OAuth. A refresh token is stored locally
(
%APPDATA%\clipdip\...) so the connection can be renewed. Token exchange happens withdiscord.com. - Please respect others' privacy: only share clips identifying other people with their awareness/consent, as also required by Setsu's Terms.
The main app's Discord Rich Presence feature (which can show "Browsing clips" on your Discord profile) is off by default and talks only to your local Discord client, no clip data is transmitted.
To find and fix bugs, both parts of Setsu can send anonymous diagnostics to a developer-operated server:
- the library app (Setsu itself), and
- the recorder (Setsu Recorder), which runs as its own background app.
They report separately and have separate opt-out switches, both shown together in Settings > About > Anonymous diagnostics. Turning one off does not turn the other off. This is designed to contain no accounts and no personal identifiers.
Identifiers used. Random UUIDs only: a per-install ID, a per-session ID, and a
per-machine key stored in your own Windows registry (HKCU\Software\ClipLib /
HKCU\Software\Clipdip, the old names, kept on purpose). The machine key is not
derived from your hardware, your Windows MachineGuid, or your username; deleting the
registry value makes you a new machine to us. The library app deliberately reuses the
recorder's install ID when one exists, so the two products' reports about the same
machine can be matched.
- What the library app sends automatically (if not disabled):
- a heartbeat roughly every 15 minutes: the random IDs, app version, uptime, a one-time hardware profile (OS build, CPU, RAM, GPU and driver, monitor resolutions, locale, time zone, and the category of the drive your clips are on (system, other, removable or network) with its free space in GB, and a small set of your feature settings (e.g. export preset, whether Discord Rich Presence is on) alongside a few coarse measurements of the app's state (library size as a bucket rather than an exact count, total library size in GB, number of tags, and whether hardware encoding, the recorder, the file watcher and a sign-in are active). It never contains your clip folder, the recorder's binary path, your API token, the sharing server URL, any tag name, or any clip name.
- error, crash and failure events: a stable error code, a category, a severity, a timestamp, and a small object of numbers and fixed keywords. Error text from the main process is included after path-scrubbing; errors from the app window send no message text at all, only the error type and top stack frames.
- aggregated performance metrics: counts and bucketed timing histograms (e.g. how long startup or an export took). No per-clip records and no timestamps of what you did.
- a daily usage rollup: one set of counts per completed day, covering how many clips you watched and for how long in total, and how many renames, trims, deletes, tag changes, exports (by format and destination), shares and imports you made. It is derived from the activity log Setsu already keeps locally for the year-end recap. That log itself is never uploaded, and the rollup carries no clip names, custom names, tag text, folder paths or share ids.
- a session end notice when the app closes.
- What the recorder sends automatically (if not disabled): heartbeats, hardware and
capture configuration, and capture/encoder/save failures. See
clipdip/TELEMETRY.md. - Note on log contents: crash reports attach a short compressed tail of the app log. The library app removes Windows user paths from that tail before sending, but it can still incidentally contain clip file names, which in turn can reveal a game name and a date/time. We do not use these to identify you, but you should be aware they can appear. Log tails are attached only to fatal errors and to specific error codes while a particular bug is being investigated, never to routine events.
- Opt-out: Settings > About > Anonymous diagnostics, one switch per product. Turning the library app's switch off takes effect immediately: no heartbeat, no events, no metrics, no daily usage rollup, and anything queued locally but not yet sent is deleted.
The full, itemised description for each product lives in TELEMETRY.md (library app) and
clipdip/TELEMETRY.md (recorder), including the complete list of what is never sent.
Manual, user-initiated uploads. One bundle covers both products: recent logs, settings files, your local activity log, captured console output, the recorder's logs and status, a crash-dump listing, system info, and the note you typed. It is far more detailed than the automatic telemetry above, which is why it is never sent on its own. (Discord tokens are deliberately excluded.) In Settings > About:
- "Save zip" writes that bundle to a folder you choose. Nothing is transmitted.
- "Export and upload" sends it to the diagnostics server. It requires you to describe the problem first, so you always know an upload is happening.
Legal basis: legitimate interest in keeping the software stable and secure , Art. 6(1)(f), balanced by the data being anonymous and the opt-out above. Manual uploads rely on your consent (Art. 6(1)(a)).
| Recipient | Purpose | What is sent | When |
|---|---|---|---|
social.setsu.gg (Setsu Sharing, operated by us) |
Accounts, clip sharing, social feed | Discord identity, uploaded video + title + tags + @mentions, banner, reactions/comments | Only after you sign in / share |
logs.yuma-homeserver.online (Diagnostics, operated by us) |
Crash/bug diagnostics for the library app | Random install/machine/session IDs, app version, hardware profile, feature settings, error events, aggregated timing metrics, log tails on fatal errors | Auto (opt-out) |
logs.yuma-homeserver.online (Diagnostics, operated by us) |
Crash/bug diagnostics for the recorder (Setsu Recorder) | Random install/machine/session IDs, app version, hardware and capture configuration, capture/encoder failures, log tails | Auto (separate opt-out) |
logs.yuma-homeserver.online (Diagnostics, operated by us) |
Manual problem reports | One bundle: logs, settings files, activity log, console output, recorder logs and status, crash-dump listing, system info, your note | Only when you press "Export and upload" |
GitHub (github.com) |
Update check & download | Generic User-Agent only | On update check/download |
Discord (discord.com) |
Login & voice-roster read | OAuth login; reads your live voice-channel membership | Only if you connect Discord |
Recording and video processing (ffmpeg/ffprobe) run entirely on your device.
- Local data stays until you delete it (your clips, sidecars, settings, logs).
- Shared clips and account data are kept while your account/shared clips exist. You can delete shared clips or ask us to delete your account and content, email us.
- Diagnostics are kept only as long as needed to investigate stability issues: error events and their log tails 90 days, manual bundles 30 days, raw heartbeats 7 days. Kept indefinitely: the daily usage rollup, and one row per install and per machine. These are pseudonymous rather than aggregate, each daily rollup is stored against the random install ID and machine key described in section 6, alongside the feature-settings snapshot, so a single install's day-by-day counts remain queryable. They contain no name, account, path or file name, and the random IDs are the only thing linking them to a machine; deleting the registry key in section 6 breaks that link going forward. IP addresses are truncated to the first two octets before storage and are never stored in full.
The Setsu sharing and diagnostics servers are operated by the developer. Depending on the hosting provider, data may be processed on servers [in/outside the EU, to be confirmed by the operator]. Where processing occurs outside the EU/EEA, we rely on appropriate safeguards under GDPR (e.g. Standard Contractual Clauses). Contact us for details.
Under the GDPR you have the right to:
- access your personal data,
- rectify inaccurate data,
- erase your data ("right to be forgotten"),
- restrict or object to processing,
- data portability,
- withdraw consent at any time (without affecting prior processing), and
- lodge a complaint with a supervisory authority (in Germany, your state Data Protection Authority).
To exercise any of these, email fabiansilbermann1@gmail.com. Note that most local data is under your direct control and can be deleted by you at any time.
- Your sharing access token is stored encrypted using your OS's secure storage (Windows
DPAPI / keychain via Electron
safeStorage). - All network communication with our servers, GitHub, and Discord uses encrypted HTTPS/TLS.
- No method is perfectly secure, but we aim to minimise the data collected in the first place.
Setsu is not directed at children. You must be at least 16 years old (the digital-consent age under German GDPR) to use the sharing service. We do not knowingly collect data from children below this age.
We may update this policy as the app evolves. Material changes will be reflected here with a new "Last updated" date. Continued use after an update means you accept the revised policy.
Questions about this policy or your data:
Email: fabiansilbermann1@gmail.com
See also the Terms of Service.