Plugins
Browse, filter, and install DeepSeek-Harness plugins.
392 plugins found
dsh-dsml-leak-guard
gogozeppeli-towa/dsh-dsml-leak-guard
Drops text blocks that carry nothing but leaked DSML tool-call tags, then rewrites the resulting empty turn into a retryable error so the host's built-in retry re-issues the request; prose that merely contains tags passes through untouched.
dsh-web-control
weichengwu/dsh-web-control
Manage a LAN or Tailscale reachable dsh web instance from inside DSH. dsh web listens on loopback by default; this adds a Settings > Web access panel that detects the machine's Tailscale and LAN addresses, starts dsh web bound to the one you pick, and shows the token URL to open on a phone or another machine. Also stops and restarts it, tails the child log, persists the bind address and port, and registers two agent tools (dsh_web_control, dsh_web_config). Zero runtime dependencies. Anyone who can reach the bound address and holds the token URL gets your dsh's full privileges (files and commands); LAN mode is plain HTTP, Tailscale is encrypted by the tunnel, and the token is single-use and changes on every start.
dsh-toolcall-id-repair
zyf2492313716-cloud/dsh-toolcall-id-repair
Stops session lockups from OpenAI-compatible gateways that stream empty tool-call ids: intercepts llm/stream chunks in memory to synthesize consistent call_<sha256> ids across tool-call-delta and block-end, plus offline session repair/verify scripts under tools/.
dsh-wait-guard
lion-li-git/dsh-wait-guard
Adds budgets to agent waiting: caps job_output waits, refuses runaway foreground commands inside run_code, breaks status-polling loops after four no-progress probes, and explains run_code deadlines and parse errors.
dsh-thinking-channel-guard
laoyingfuhuole/dsh-thinking-channel-guard
Moves reasoning that a provider leaks into the visible text channel back into the reasoning channel, and emits a text block when a whole turn of answer text lands only in the reasoning channel. Catches three shapes: `<thinking>`-tagged text, text that continues a reasoning tail which ended mid-token, and an answer-only reasoning turn that stops. Healthy streams pass through unchanged, block order is preserved, and each of the four rules can be switched off in one JSON config.
dsh-ml-workbench
shinerain1/dsh-ml-workbench
Track, branch, compare and reproduce local CSV classification and regression experiments through DSH tools backed by MLflow; requires a separately installed Python backend. By default training runs as detached local worker processes that keep running after DSH exits (executionMode dsh stops them with DSH); running your own Python preprocessing or estimator code needs allowCustomCode, which is off by default; the plugin makes no network requests.
dsh-browser-ops
dcd887/dsh-browser-ops
Self-managed persistent browser for DSH automation — launches its own Chrome/Edge profile, navigates, switches tabs, snapshots with semantic refs, acts by ref, evaluates JS, screenshots, and uploads files. Does not bypass CAPTCHAs by design.
dsh-doom-loop-guard
dixiewithheart/dsh-doom-loop-guard
Detects repetitive loops in an agent's reasoning output and records, nudges, or cuts them, with a history report. Defaults to a nudge (a reminder injected into the next request); cutting and auto-resuming is opt-in, and should not be combined with another plugin's cutting such as dsh-loop-rescue. Needs DSH 0.2.1+ and Node.js 22.15+.
dsh-redo-live
tanxiaolong/dsh-redo-live
Edit and resend or regenerate a message you already sent, including while its answer is still streaming: the plugin stops the running turn through the official Session.cancel, waits for the session to go idle (8s cap), then rewinds that message in place the same way dsh-redo does. Command is /relive so it can coexist with dsh-redo.
dsh-delete-session-menu
ba1ying2/dsh-delete-session-menu
Adds "Delete conversation" (zh/en) to the sidebar conversation menu. Confirming removes the durable session log under <DSH_HOME>/sessions, drops the id from the workspace registry (sessionIds, pin, archive), detaches a live session before its files go away, and emits api-session/removed so every connected client drops the row.
dsh-task-budget
andromder/dsh-task-budget
Per-task cost dashboard and budget guard that prices each model call at peak or off-peak rates, turns the remaining budget into a per-turn work envelope in the system prompt, and stops the task at its cap until the user replies. The default cap is ¥0.5 per task.
dsh-big-coming
sumcam-l/dsh-big-coming
Watches the official DeepSeek models for changes: every cycle it reads the model catalog and sends one minimal completion per watched model, then compares the system fingerprint, the served model name, the response envelope shape, the catalog metadata and the token count of a fixed prompt. When something moves, it pops a random meme with a table of exactly what changed and plays a blast. Needs a DeepSeek API key (kept in the credential store); by default it runs every 60 minutes and spends tokens on each watched model.
dsh-ps-preflight
dingchenhui0618-arch/dsh-ps-preflight
Grades every PowerShell command an agent is about to run before it runs: deletion, overwrite, network download and credential access are four required classes, plus dynamic execution, system and privilege changes. Verdicts are a pure deterministic rule function — no model call, no network, sub-millisecond — with two escalation rules (download+execute, credential+exfiltrate) that jump straight to high. High risk stops the tool call for a second confirmation (in-panel queue or the native approval), deny or timeout means it never runs; medium is logged but never blocked. Every command and its verdict, including the redacted command text, is appended to a local JSONL audit log. Quoted danger strings are treated as data, -WhatIf is respected, -EncodedCommand is decoded and analysed again. Host + client halves, zero runtime dependencies, 55 rules, 136 tests.
dsh-dev-launcher
zhangkaiqiu/dsh-dev-launcher
Saves per-workspace dev commands behind a header button, starts and stops them as process trees, and exposes four tools so the agent can list, start and stop those saved commands and read their logs (it cannot add or change commands).
ascend-npu-profiling-tuning-skill (dsh-ascend-perf)
newpage1/ascend-npu-profiling-tuning-skill
Ascend NPU fleet monitor plus controlled vLLM container deployment for DeepSeek Harness. Environments come from envs.json; npu-smi is collected locally or over SSH with one connection per environment. The sidebar shows idle/partial/busy occupancy, processes, and per-card/per-chip details, and lets you preview or start a vLLM container on selected idle NPU chips. Preview is read-only; real start requires deploy.allowRealExecution plus a confirmation token, only touches dsh-apf-svc-* containers, binds ports to 127.0.0.1, and never stops other users' containers.
dsh-session-alerts
qyf1999/dsh-session-alerts
Badges the installed app icon with the number of finished, unread sessions, and pops an in-app badge or a Windows notification when a session fails or waits for your answer; each state has its own chime.
dsh-clamawake
zhangweijian97/dsh-clamawake
Keeps the Mac awake for as long as the DSH host is running, whether or not an agent is working: starts ClamAwake's virtual display (mirroring the main screen) plus caffeinate -d, stops both when DSH exits and restarts the engine if it crashes. Self-contained (bundles the arm64 vd-engine, GPL-3.0). Requires macOS on Apple Silicon.
dsh-computer-use
mount9527/dsh-computer-use
Windows desktop operation layer for DSH: 17 tools for window listing, screenshot plus OCR, UI Automation, keyboard and mouse, file operations and CDP browser control. Non-intrusive by default (UI Automation patterns and window messages, never the physical pointer), irreversible steps go through a native confirmation dialog, Esc or an on-page button stops a running task, and a global lease keeps two sessions from driving the desktop at once.
dsh-arya-computer-use
konnectsuite/dsh-arya-computer-use
Drives Windows apps through accessibility controls instead of pixels, with a takeover overlay that frames the window being driven, marks each action with a ghost cursor, and stops on Esc; a browser lane adds sign-in profiles, saved state, and attach-to-my-Chrome. Windows only, and DSH must be run as Administrator for the desktop lane.
dsh-session-rename
mkpoli/dsh-session-rename
Adds one agent tool that renames the current session, keeps the title across resume and paging, and stops automatic titles for it.
dsh-token-curve
stao-gee/dsh-token-curve
Sidebar token and cost curve for the current conversation plus a local ledger across all sessions: monthly/weekly rollups, per-turn detail, Excel export with print-ready A4 layout, and a live account balance. Reads session logs only, prices every message by its own peak/off-peak timestamp, and drops fork-inherited history so it is never billed twice. Desktop app only, verified on Windows; the live balance needs a signed-in DSH account.
dsh-store-ops-expert
igdwei/dsh-store-ops-expert
Restaurant store operations expert: reconcile a POS export, locate the gap, split volume vs price, measure discount efficiency, and open a single-file local-data dashboard.
dsh-movie
hl3207/dsh-movie
Adds a left-sidebar panel to the DSH Web GUI: one ranked board of the films and series released in the past 21 days or due within the next 31 days, taken from JustWatch for the US, five European countries, Japan and Korea and from Douban for China. Each row carries a poster (or a drawn placeholder when the poster CDN will not load), a synopsis of five sentences or fewer, Douban and Rotten Tomatoes ratings and popular Douban comments. A search box filters the board as you type and asks Douban, TVMaze and the iTunes storefront about a title, merging the three into one record and listing the matches already on the board separately. Synopses, comments and titles are translated through the host translator service, so the Chinese/English switch re-renders the board. No API key and no account.
dsh-codex-model-picker
caidows/dsh-codex-model-picker
Codex-style composer model picker: a card showing the current model and a draggable slider that sets reasoning effort, with one stop per level the model advertises and an Off stop when the provider supports it. The slider snaps only to whole stops and updates the effort label while dragging; selecting a model or effort still goes through the session model directory, so it stays in sync with the rest of the composer.