Adds a local-only six-seat Texas Hold'em table to the DSH Web UI with full betting streets, side pots, persistent Token stacks, and agent-status display.
Install
# from a prebuilt release tarball
dsh plugin --profile web add "https://github.com/leeclouddragon/dsh-all-in/releases/download/v0.1.1/dsh-all-in-0.1.1.tgz"
# from GitHub (first run asks for allowBuilds approval — follow the hint, retry)
dsh plugin --profile web add github:leeclouddragon/dsh-all-in
Any plugin you install runs third-party code with your own permissions — it can read your files, use your credentials, and reach the network, and tool approvals don’t sandbox it. GitHub-sourced plugins also run build scripts at install time — pnpm blocks those until you allow them, so an install can stop with ERR_PNPM_GIT_DEP_PREPARE_NOT_ALLOWED or ERR_PNPM_IGNORED_BUILDS; dsh prints the exact key to add under allowBuilds in your profile’s pnpm-workspace.yaml, and the install works on the next run. Allowing a build is a trust decision: only install sources you trust, and pin a commit (github:owner/repo#sha).
README
Your agent is thinking. You're all in.
Play a fast, local-only six-max Texas Hold'em hand while DeepSeek Harness is thinking.

dsh-all-in adds an All In action to the Harness sidebar and opens a full-screen poker table through the supported shell.overlay slot. The active agent keeps running underneath it. The game never writes to the conversation, calls a model, or uses a network connection.
Entertainment only. Tokens have no monetary value. There are no deposits, withdrawals, purchases, accounts, or multiplayer wagering.
What works
- Six seats: one human and five local bots
- Monte Carlo bot equity with pot-odds, opponent-count, position, stack-to-pot, and personality-aware decisions
- Persistent Casual, Competitive, Shark-table, and experimental solver-style GTO opponent levels, adjustable from the table header
- Occasional action-aware table talk written for each bot personality
- Pre-flop, flop, turn, river, showdown
- Fold, check, call, pot-sized raise, and all-in
- Custom raise-to sizing with Min, 1/2-pot, 3/4-pot, Pot, Max, slider, and direct amount entry
- 5M Token starting stacks with 25K/50K blinds
- Seven-card hand evaluation and split pots
- Side-pot settlement for unequal all-in contributions
- Persistent local Token stacks across page reloads
- Live Agent thinking / Agent idle status
- English and Simplified Chinese UI
- Responsive desktop and narrow-window layouts
Install
From the prebuilt GitHub Release:
dsh plugin --profile web add https://github.com/leeclouddragon/dsh-all-in/releases/download/v0.1.1/dsh-all-in-0.1.1.tgz
dsh --profile web
From npm after the first npm release:
dsh plugin --profile web add dsh-all-in
dsh --profile web
From a local checkout:
git clone https://github.com/leeclouddragon/dsh-all-in.git
cd dsh-all-in
pnpm install
pnpm check
dsh plugin --profile web add "$PWD"
dsh --profile web
For a GitHub source install, DeepSeek Harness uses pnpm's git dependency flow. The package has a prepare script that builds lib/; pnpm 10+ requires users to allow that build explicitly. Publishing prebuilt artifacts to npm avoids that prompt.
Development
pnpm install
pnpm check
The package has two halves:
lib/index.js: a no-op host plugin that activates the bundlelib/client.js: a browser bundle registeringsidebar.footer.actionandshell.overlay
The poker engine is independent of React and exported as dsh-all-in/engine. Bot decisions estimate equity from only their own hole cards and the public board, so simulations never inspect an opponent's hidden cards or the real deck order.
Why this is a separate repository
DeepSeek Harness is currently in developer preview and its contribution guide says external pull requests are not accepted yet. The official ecosystem path is an independent plugin repository tagged with the GitHub topic dsh-plugin.
License
MIT
Links
More in this category
PC2005-cloud/dsh-pet#dsh-pet★ 1069
Desktop pet for the DSH Web UI with 25 transparent animations, screen wandering, click reactions and drag, plus a reproducible asset-generation pipeline.
Nagi-ovo/dsh-ads★ 647
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.
vlln/whale-girl★ 345
Desktop pet (QQ-pet style): floats in the corner, draggable, feedable, playable.
a86582751/dsh-nexttavern★ 146
Roleplay workspace for DeepSeek Harness: import SillyTavern/TauriTavern character cards or author one from scratch through an interactive flow; turn a long TXT novel into a playable card with either an intensive or a coarse reading pass; the main agent consults worldbooks on demand, and prose is read in a dedicated tavern reader tab; explore multiple worldlines inside one conversation (regenerate, edit-and-send, explicit clone); scoped style presets with 16 built-in styles, a character agent cluster, a standalone decision card, and card or novel export; long stories hold together through a frozen settings prefix, a hard-cut context window, traceable director notes and keyword/semantic/hybrid recall, with embeddings either online or fully local. Targets Harness 0.1.2-alpha.3 and applies its documented compatibility patches.
yyh-001/dsh-meme★ 145
Chat meme stickers: text-only send, mood auto-send, QQ/WeChat-style picker, auto-learn, custom packs.
Sutera-Diffusus/dsh-whale-musume★ 112
A whale-girl desktop pet for the DeepSeek Harness Web UI: pat-to-raise growth, work-state poses per tool type, 494 dialogue lines, 30 achievements, a growth diary, drag physics, theme sync and an accessibility mode. Chat-keyword sensing and API-balance care ship off and are opt-in; an optional weather mood calls Open-Meteo only after you name a city. No telemetry.
Community comments
Comments are public GitHub Discussions. Loading them connects to GitHub and Giscus; a GitHub account is required to post.