Plugins
Browse, filter, and install DeepSeek-Harness plugins.
160 plugins found
dsh-pet-ronaldo
stellum-waq/dsh-pet-ronaldo
Cristiano Ronaldo desktop pet for DeepSeek Harness: a chibi Portugal number 7 mascot in a bottom-right overlay whose animation follows the agent state — dribbling while tools run, thinking during a turn, waiting at approvals, diving on errors — and jumping in a SIU celebration with a host-process sound when a turn completes; it can be dragged, looks at the cursor on hover, dives on a triple click, and a settings panel manages several pets and imports custom spritesheets. It also runs as a borderless, always-on-top native Windows window that keeps living after the browser is closed: hovering shows the current workspace and conversation title, double-clicking opens the Harness page in Edge, and dragging, mouse-wheel resizing and a right-click menu work there too, sharing position, size and behaviour with the web panel. Pet sheets may use any cell resolution and scale preserving aspect ratio (atlas.scaling: smooth | pixelated). v2.2 adds a video route — feed it a clip (green screen or ordinary footage) and its real motion becomes the pet animations, via automatic backdrop sampling, chroma keying with spill suppression, motion-based segment splitting and union-bbox alignment; already-extracted PNG frames work too, with no video decoder installed, and the clip audio can be cut into per-action sounds. Pets can also be published to the community gallery under the DPSL-1.0 sharing declaration, and installed from it in one click. v2.3 adapts it to DSH 0.2 desktop clients — the celebration sound now falls back to spawning PowerShell when the host no longer provides the shell service — and makes a single click play a short, configurable taunt animation (`interactions.clickAnim`).
SpecWave
cyning12/specwave
ICVO coding-standards kit for DeepSeek Harness (formerly dsh-coding-kit): a bundle plugin that registers tools on load and injects the standards only when apply_coding_standards is called, plus a P0 gate CLI (npx spec-wave) with G1-G7 process commands, skills install, host apply, and IDE-block refresh.
dsh-agent-sync
kuaiyukuaikuai/dsh-agent-sync
Scans 20+ AI agents on this machine (Codex, Claude Code, cc-switch, Hermes, opencode, Gemini, Grok, Kimi, CodeBuddy, Trae, OpenClaw, Qoder, WorkBuddy, Zcode, Cursor, Windsurf, Cline, Roo Code, Qwen Code, ...) and one-click syncs, manages and adds their MCP servers and skills into DSH — global or per-workspace.
dsh-rss
stardustlc666/dsh-rss
RSS/Atom subscription tools (list/add/remove/fetch/check/search/opml_import/opml_export/health) with cross-feed keyword search, cursor-based incremental fetch and OPML import/export, persisted to settings. A step-by-step first-run guide (3 steps) now opens inside its settings panel.
dsh-agent-session-import
howtomakeaname/dsh-agent-session-import
Imports local conversation history from Cursor, Codex, Claude Code, ZCode and Kimi into DeepSeek Harness sessions, with a read-only scan, a Faithful mode that keeps tool calls, and a Resume-safe text mode.
agent-immune
meat-daemon-404/agent-immune
Self-evolving pitfall immunity for coding agents (Claude Code, Codex, Cursor, ZCode, DSH): capture failed-then-fixed tool episodes as SQLite+FTS5 antibodies and inject them before the agent repeats the mistake. Zero runtime dependencies.
dsh-skill-mover
mjylfz/dsh-skill-mover
Migrates skills from other agents (Cursor, Claude Code, Codex, OpenCode, and more) into DSH: scans 14 agent platforms plus the shared ~/.agents skills layer, detects duplicate names across sources, lets you pick the best source per skill, copies them into the DSH skills directory, and rolls back previously migrated skills.
back4app-plugin
back4app/back4app-plugin
Skill that deploys a folder to a public HTTPS address on Back4app with the Back4app CLI, on a temporary account with a claim link to keep the web app.
oh-my-agents-above-all
gulagala001/oh-my-agents-above-all
Codex, Grok Build, Cursor, Pi and ZCode presets with native conversations, code review, document conversion and workflow controls.
dsh-native-browser
longmiaoo/dsh-native-browser
Browser plugin for DSH (浏览器): Codex-style native browser control in your signed-in Chrome, with smooth virtual-cursor feedback, semantic page reads, form filling, screenshots and handoff; setup from the Browser sidebar. Requires macOS + Google Chrome (Chrome 125+) and a one-time unpacked load of the bundled MV3 extension; Edge, Windows and complete Linux installation are experimental.
dsh-skin
changan0617/dsh-skin
Homepage-style chat background skin for DeepSeek Harness, ported line by line from deepseek.com/harness: WebGL fluid background (5-colour noise field, flow-field brush, light, bloom, vignette, grain), white 90px dot-line grid with mouse perturbation and spring return, and a digital point-cloud whale that assembles from particles, breathes, swims, is pushed radially by the cursor and brightened by a light beam. Dark theme only; disabled automatically under reduced-motion.
dsh-connect-cursor
greatyao/dsh-connect-cursor
Connects a locally-signed-in Cursor desktop app to dsh as a native LlmAdapter — reads the Cursor login token from state.vscdb or auth.json, drives the private agent.v1.AgentService/Run HTTP/2 bidi stream, and translates InteractionUpdate deltas into dsh StreamChunks. No loopback proxy. macOS, Windows and Linux.
dsh-cursor-dev
dclef/dsh-cursor-dev
Runs the local Cursor CLI as a child coding agent from DSH - a `cursor_code` tool plus the native one-shot `cursor_subagent` tool, one mirrored DSH sub-session per run, per-task model choice, resumable conversations, a step timeline with real file diffs, and a usage, model and session panel.
dsh-selection-quote
xt765/dsh-selection-quote
Cursor and GitHub PR Review style selection toolbar: quote selected text to chat (Cmd+L), write inline comments (Shift+Cmd+L), and batch-insert from a composer dock.
dsh-smooth-cursor-patched
thinkofrain1213/dsh-smooth-cursor-patched
Comet-mode animated caret for the DSH Web composer: a glowing trail glides to the text position, with configurable accent color, thickness, trail, and a 500 ms idle blink. Extends the surface to ask-question card inputs, and fixes the caret vanishing on an empty composer or first focus, forward drag-selection not following the pointer, soft line breaks measuring against the previous line, and the caret painting outside the composer when long input scrolls. Maintained fork of Lacquervii/smooth-cursor.
spritely
wang-junjian/spritely
Spritely — a floating work-state mascot for the dsh web UI: it animates with the agent's live activity, tracks the cursor, and offers switchable characters, draggable placement, and customizable backgrounds.
smooth-cursor
lacquervii/smooth-cursor
Smooth comet-mode animated caret for the DSH Web composer: a glowing comet trail replaces the native blinking caret and glides across the input as you type, with configurable accent color, thickness, and trail, IME-aware for CJK input.
dsh-git-management
et921/dsh-git-management
Git panel for the DSH right sidebar: staging, commits, branch switching, fetch/pull/push, worktrees, a Cursor-style commit graph and GitHub pull requests (the GitHub card needs a GitHub token). It drives the system git binary, so the panel and the terminal always agree.
dsh-local-trim
leonlee114514/dsh-local-trim
One-step local model onboarding plus a light harness for DSH. On load it probes loopback and private-network inference endpoints (llama.cpp, Ollama, vLLM, LM Studio), reads each model's context window and output budget from the backend, and registers the provider through settings, so a local model becomes selectable without hand-writing config. It then trims only that provider: masks the tool list through tools.restrict, drops named system-prompt sections and memory contexts, swaps the injected instruction block for a short one while preserving its first 80 and last 96 characters, and truncates tool-schema descriptions without touching structure or required fields. Cloud providers are returned as the same assembly object, untouched. The injected block is re-read from disk on every assembly, so editing the file takes effect on the next request with no reload. The panel in the plugin detail page is a live operations view: it shows whether the takeover is active and by which hook, how many characters the last assembly was trimmed from and to, and which providers are online (probed concurrently, with a real one-shot completion reporting latency and reply text per provider). If a local backend is not running, the panel offers a one-click launch: it probes the port first (never starting a duplicate), then starts the matching server (ollama serve, or a custom command you supply) detached and polls until the endpoint answers, clamping the wait to 3-600 seconds, reaping the child it just spawned if it never answers, and naming the paths it searched when it cannot find the binary. A stop button kills backends this plugin started (SIGTERM, then SIGKILL) and refuses to touch a server somebody else launched. When a model has just been downloaded and no backend exists yet, the panel can also scan the machine for .gguf files with hard caps (300 models, 20000 directory entries, Downloads only one level deep), start llama-server on the chosen one (skipping the .mmproj. vision projection and reusing a server that already answers), register the provider under the model id the backend reports, and run one real completion to prove it works. Before spawning anything it does a fit check that counts only the part of the file the GPU cannot hold: hostBytes = fileSize - VRAM, compared against max(total RAM - 2GB, free RAM). A file far larger than RAM is refused outright; a heavy host-side remainder is refused unless you confirm; the same remainder under mmap is only a warning, because the weights land in reclaimable page cache. A load-mode selector (auto / mmap / none / mlock / mmap+mlock) maps to llama.cpp --load-mode, and the default is whatever the engine picks. From 0.4.2 everything above is derived from the platform instead of assumed: home and data directories come from os.homedir() and the OS conventions (LOCALAPPDATA on Windows, ~/Library/Application Support on macOS, XDG_DATA_HOME elsewhere); backend binaries are looked up in the official install locations of each OS and can be pinned with OLLAMA_BIN, LLAMA_SERVER_BIN, VLLM_BIN or LMSTUDIO_BIN; model discovery walks DSH_LOCAL_MODELS_DIR, ~/models, ~/llama.cpp/models, ~/.cache/llama.cpp and the LM Studio data directory, with the WorkBuddy layout demoted to one candidate among them. VRAM is asked from nvidia-smi, then rocm-smi, then sysctl hw.memsize on Apple Silicon (flagged unified, so the same bytes are not charged twice), and any flip to a unified-memory budget is handled explicitly. Before building a command line the plugin runs llama-server --help once and keeps only the flags that build actually has, so older engines get --n-gpu-layers 99 and no KV quantization while newer ones get auto plus q8_0; anything it cannot probe is passed through rather than dropped. LM Studio is started with lms server start --port N rather than being lumped in with the engines that need a model argument, and vLLM gets its own --model example instead of llama.cpp syntax. Windows, macOS and Linux are all exercised in CI, including a windows-latest and a macos-latest job, because a cross-platform claim that has only ever run on one OS is not a claim. Two of those suites exist specifically to catch this machine's blind spots: one replays the whole suite with process.platform rewritten to each OS, and one deletes every advantage this development box has — no nvidia-smi, no rocm-smi, no sysctl, an empty PATH, a home directory with no models — and requires graceful degradation rather than a crash. That last one caught a real class of bug: several assertions hardcoded 'lms.exe', which cannot pass where the executable has no suffix. Measured on the reference machine: a 16.9GB 27.2B MoE with 14.5GB of VRAM holds a 285-691MB working set and sustains 76 tok/s after a 1000-2000 token warm-up, so file size alone is not a fit signal. From 0.4.3 the scanned models are shown as a table rather than a dropdown (one row per gguf, with its size and any fit warning), so you can see what is on disk before choosing. From 0.4.4 the delete button on the provider table removes the provider entry, not the weights file on disk. That distinction matters: what people actually want to clear out is the dead entry - one pointing at a port that never answered, one registered twice for the same port, one whose model file was deleted long ago - not a 15.8GB gguf. Deleting a config is undone by importing again; deleting a gguf is not. The removal goes through settings.mutate with an unset op on \['providers', id], because update merges layers: writing null there leaves a null provider key that the loader reads as a broken config rather than an absent entry, and only mutate/unset is implemented as Reflect.deleteProperty. The result is read back afterwards, because unset on a key that comes from a lower layer (the profile cordis.patch.yml or a home patch) does not delete it - it restores the inherited value, so the call succeeds while the entry stays. When the entry is still there the panel says so and points at the file to edit instead of reporting success, and a DSH whose settings service predates mutate() is told plainly rather than silently ignored. Deleting takes two clicks and clearing the confirmation is as easy as clicking another row; on success the online and baseURL caches are cleared and the tool surface is recomputed, since the provider list changed. The gguf itself is never touched by this plugin - the scanned-model table is display and selection only.VRAM reporting is driver-dependent. From 0.4.5 model discovery also reads Ollama's own storage: its blobs are named sha256-… with no extension, so extension matching could never list them. Weights are now accepted by extension OR by GGUF magic header (only dot-free files of at least 8MB, capped at 600 header reads per scan; obviously-not-weights extensions like .json/.md/.safetensors/.part are never opened). Ollama models are resolved from their manifests: the model layer is picked by mediaType application/vnd.ollama.image.model (never adapter/mmproj layers), the human-readable name is restored from the manifest path (…/qwen3/8b becomes qwen3:8b, …/latest stays bare), tags sharing one blob merge into a single row, and the metadata kv supplies name, quantization and context length. Adopting an Ollama model does not start a duplicate llama-server over the same weights: it confirms ollama serve answers on 11434 (launching it if not) and registers the provider under the tag the endpoint itself reports — the blob's sha256 name would never match anything, since 11434 lists qwen3:14b, not sha256-a8cc…. Two real-machine-only traps fixed on the way: existsSync('<file>/') returns true on Windows, so manifest-vs-directory branching now goes by parseable content instead of trailing-slash semantics; and the quantization label table had been copied from a stale ggml enum generation, labelling ggml_type 15 as Q5_K_S while ollama show reports Q4_K_M. Same-blob dedup is by (dev,ino), keeping whichever entry has a name. Verified on the reference machine: the scan goes from 1 row to 4 — qwen3:14b (8.64G) and qwen3:8b (4.87G) now appear with Q4_K_M badges — in 4ms; npm run verify exits 0 (launcher 57, portable 35, panel-host green including the linux/darwin simulated-platform runs); the new panel-host suite builds a real miniature Ollama store in tmpdir and points OLLAMA_MODELS at it, so results no longer depend on whether the test machine happens to have Ollama installed. From 0.4.6 the panel shows each model its real name instead of its file name, because some download tooling stores every weight under one generic file name, leaving the real name only inside the GGUF header. The scan now reads each file GGUF KV section - not a fixed-size head buffer, because tokenizer.ggml.tokens (hundreds of thousands of strings on a large MoE model) and the chat template push general.file_type past any window; and GGUF string arrays are interleaved (u64 length plus payload per element), so treating them as a length table walks the cursor off the rails and produces astronomically large reads. Reading and skipping are accounted separately: key names and types are always read, string values over 256KB and string arrays are only walked past without copying, non-string arrays over 64 elements are skipped outright, and a cursor past the file size stops the read, because an out-of-range position means the cursor is wrong rather than that a limit should be raised. The displayed name goes Ollama tag, then general.basename plus size_label, then general.name, then the file name, with the full string kept for the hover text since it usually carries internal build numbers; mmproj files are now identified by general.type as well as by name. A short read (truncated file) no longer feeds half a value into the parser. On the reference machine, reading one file header parses 44 keys in about 16ms, and a whole scan takes about 31ms. From 0.4.7 the smoke test no longer fails the first time and passes the second. The old code gave both request attempts one shared AbortController and one shared 30-second timer, so the total budget covered both rounds: a cold load of a large model burned it in round one, the abort left the signal permanently aborted, and the second round (the Ollama-style reasoning_effort: none retry) could not even be sent - which is exactly why the first click returned This operation was aborted and the second succeeded. Now each attempt gets its own controller and timer, a timeout retries once automatically (retryOnTimeout, on by default) because a first-request timeout almost always means the weights are still being read from disk, and non-timeout errors (refused connection, HTTP error) are not retried at all. Timeouts report themselves: the message says the wait exceeded smokeTimeout, whether a retry already happened, and points at the new smokeTimeout setting (seconds, default 60, clamped to 5-600) - port responding is not the same as tokens coming out, so it is a separate knob from launchTimeout. Unrelated: an example naming one specific model was dropped from the panel copy, since not every machine has that model downloaded, and the explanation now describes the mechanism instead. From 0.4.8 adopting a plain .gguf no longer registers an unrelated Ollama model. The port field used to follow the last click rather than the file: selecting an Ollama entry moved it to 11434, selecting a plain .gguf did not move it back, so a 11434 left over from an earlier click stayed. The server accepted it as given, probed the port, found Ollama answering and borrowed it (the skipped-launch log line about the port already answering is this path), and then the model id - derived from the file name - could not be matched against Ollama's /v1/models, where the fallback was to grab the first entry in the list. Embedding models tend to sort first, so the adopt registered a model the user never selected. Now the port follows the file: a plain .gguf landing on another engine's port (ollama 11434, vLLM 8000, LM Studio 1234) is moved to 8090 and the panel says so. Only ports belonging to a known engine are overridden - a hand-typed 9999 aimed at your own llama-server is left alone. The model-id fallback no longer guesses either: it takes the only entry when there is exactly one (a single-model backend may legitimately report a different id than the file name) and otherwise reports an error listing the ids it did see, rather than registering a wrong provider. From 0.4.9 the panel can no longer vanish. It used to be possible for the whole section to disappear from the plugin detail page while the host stayed up, the HTTP bridge kept answering, and no crash log appeared. The cause was a scope error: a helper was declared inside one component and referenced from another, which is valid JavaScript and passes a syntax check, so it only threw at render time - when the user pressed the scan button and React ran the state updater. The host wraps every slot entry in an error boundary whose fallback is an empty anchor, so the throw was contained silently: the panel was not uninstalled or crashed, it was replaced by an empty div. The helpers now live in the module scope both components share. The client half also gained its own test - a minimal React stand-in loads the real panel module, really calls the components, really runs the state updater at render time, and presses the scan button, so this class of error can no longer reach a release.
dsh-load-agent-rules
acus9345/dsh-load-agent-rules
DeepSeek Harness plugin that loads the rule files other coding agents use (Cursor .mdc rules, Copilot instructions, Windsurf, Cline, Kiro, Roo, GEMINI.md) with their glob scoping: always-on rules up front, glob rules when matching files are touched, the r
dsh-takeover
ikeasven69/dsh-takeover
Session takeover: read-only pull local sessions from nine external agents (Claude Code, Codex, OpenCode, ZCode, Pi, WorkBuddy, Cursor, Grok, Gemini CLI) into distilled six-section handoff cards — one click on the Relay chip takes over. /handoff checks the current session into the handoff: 1 open-protocol inbox (~/.handoff, consume-once with machine-signature tamper detection); any agent picks up on start and continues. Memory files ride with cards; cross-machine relay is one env var away.
air-pixel-liquid-glass
air-pixel/air-pixel-liquid-glass
A configurable Liquid Glass theme for DeepSeek Harness with translucent surfaces, wallpaper, desktop transparency, window blur, custom cursors, click ripples, and per-side conversation contrast controls.
dsh-select-translate
shan5246992zhui/dsh-select-translate
Select text in any Windows app and read the translation in an always-on-top card next to the cursor: streaming text follows the newest line, up to 15000 characters per request. Windows only.
MACKORN-hydraulic-cone-crusher
leifdai/mackorn-hydraulic-cone-crusher
Cone crusher selection and crushing-plant design for MACKORN NH/NS single-cylinder hydraulic cone crushers: requirement-form intake, equipment selection, plant configuration, proposal generation, product-size simulation and field-data calibration. 21 tools, 6 skills, an MCP server, no runtime dependencies.