vlln/whale-girl
Desktop pet (QQ-pet style): floats in the corner, draggable, feedable, playable.
Install
via npm
Resolves a published tarball rather than cloning the repository, and installs without any extra setup. Swap `web` for your profile name if you run another one.
via GitHub · npm package
Installing from GitHub runs the project's build script, which pnpm blocks until you allowlist it — run the command once and pnpm prints the exact key to add under `allowBuilds` in ~/.dsh/profiles/web/pnpm-workspace.yaml.
What happened when we ran it
Installed cleanly when we ran it
Every command here is run in a throwaway container against a clean profile, and the result is whatever the harness recorded — not a guess from the source. Last run 13d ago.
Show it in your README
For maintainers: the badge serves this listing's latest sandbox verdict, so a re-run updates it on its own — and it links readers to the full result here.
Due diligence
What it is — A QQ-pet style floating desktop pet supporting drag, feed and play.
Who it is for — Users conducting sessions and tasks in the DSH web GUI who need a pet companion for interaction. Users not using the DSH Web GUI don't need this, because it is built for the DSH Web GUI.
Watch out — Sandbox test passed after installation in fresh profile and registration into the profile. No obvious issues found.
The verdict — I would install it because it converts task completion, sessions and companionship time into pet levels, titles and memories.
Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.
whale-girl is a DeepSeek Harness plugin that adds a draggable desktop pet to the DSH web interface.
It is delivered as an official bundle through the web profile and appears in the lower-right corner after the web interface restarts. The pet reacts to harness activity: it welcomes new sessions, accompanies thinking and running sessions, waits for approval, celebrates completed tasks and rounds, and shows error or disappointed states after failures. It also tracks accumulated experience from tasks, sessions and active companion time, exposing levels, titles and recent shared memories. Clicking the pet opens feeding and play actions; hovering shows its status, and dragging moves it. The host settings section whale-girl: controls the web rendering, size, opacity, walking and sleep timing.
This is intended for developers who want an in-interface companion and a visible, persistent progress layer around agent work. It is the wrong choice when an unobtrusive workspace is required, or when a separate desktop pet is already running: the optional desktop/ companion renders the same character outside the web interface and can otherwise result in two pets. That companion is not installed by dsh plugin; it uses a Node engine with a Tauri rendering shell and consumes the plugin’s public endpoints. The web pet is initially hidden during onboarding.
The web pet is hidden during initial configuration and onboarding. Once visible, it uses activity from the harness interface to change state:
| Event | State |
|---|---|
| New session | welcome |
| Session running or thinking | think, occasionally working |
| Waiting for approval | wait |
| Task, upgrade, title or round completion | celebrate |
| Task failure or request error | error, then disappointed |
No activity for at least 60000 ms |
sleep |
| Interaction while sleeping | wake |
| Feeding or playing | eat or play, then joy |
It also cycles through idle and walk, and shows random blinking or turning while idle. The state machine’s full priorities and transitions are documented in docs/state-machine.md.
Settings are read from the host settings UI or the whale-girl: section in <dshHome>/settings.yaml. Changes take effect without a restart.
| Key | Default | Meaning |
|---|---|---|
enabled |
true |
Enables web rendering |
size |
110 |
Pet size in px; accepted range is 64–160 |
opacity |
1 |
Normal opacity; accepted range is 0.2–1 |
walk.enabled |
true |
Enables periodic walking |
sleepAfterMs |
60000 |
Idle time before sleeping |
The experience and title semantics are not configurable. The character can be changed from the character menu or through localStorage key whale-girl:character.
dsh plugin --profile web update whale-girl updates the installed plugin; a web restart is required for the bundle change to take effect. The separate desktop/ application is not installed by dsh plugin. From that directory, npm install prepares the application, while cd src-tauri && cargo run starts the recommended Tauri shell. npm run start:headless starts its headless mode.
The companion uses the plugin endpoints /state, /events, /presence, /interact, /config and /assets. When it is active, the web pet hides through the presence contract; after exit or a 45s presence TTL expiry, the web pet returns.
Written from the project's own documentation and kept in sync with it. Where the two disagree, the source is authoritative — read the README on GitHub
Same category
zhu1090093659
A whale-girl companion for the dsh web GUI: animations follow the agent state, with pat interactions and a feeding mechanic that raises intimacy from a calf to deep-sea bonds.
Installed cleanly when we ran it
PC2005-cloud
Desktop pet for the DSH Web UI with 25 transparent animations, screen wandering, click reactions and drag, plus a reproducible asset-generation pipeline.
Installed cleanly when we ran it
Nagi-ovo
Parody ads in 2005-Chinese-web style: sidebar banners, in-chat feeds, corner popups, and a close button whose hit area is smaller than it looks. All fictional.
Installed cleanly when we ran it
QCYTSN
Desktop-native BigFish companion for DeepSeek Harness — real Agent status, always on top on Windows.
Installed cleanly when we ran it
yyh-001
Meme search, the fun way: describe the vibe, and the agent finds and sends a real meme that actually fits.
Installed cleanly when we ran it
yyh-001
Chat meme stickers: text-only send, mood auto-send, QQ/WeChat-style picker, auto-learn, custom packs.
Installed cleanly when we ran it
