Plugins

Screenshot editor that lands the crop straight in the composer draft.

A video tool for DeepSeek Harness: watch live streams, videos and podcasts with real-time subtitles in any language, then clip and download

HeiXia, a cat companion for DeepSeek Harness with actions based on photos and video, session-state reactions, chat, and balance display. A fork of PC2005-cloud/dsh-pet that uses the same plugin id, so it replaces the original — do not install both together.

Evidence-gated teaching-material workbench with a Web UI: page model, per-page handout and slide editing, PAEG Chinese language checks, layered gates, checks center, version history, handout/deck/storyboard/Manim-script export and rights audit.

Intervals.icu-only endurance-coaching skill bundle that turns Intervals.icu training data (activities, intervals, streams, training load CTL/ATL/TSB/rampRate, HRV, resting heart rate, sleep in-bed hours, eFTP, power curves) into concrete training decisions: readiness gating, VDOT pace and heart-rate zones, periodized weekly plans written back to the Intervals.icu calendar as events after confirmation, session reviews, race-day pacing and fueling, and weather-adjusted workouts. The bundle registers the endurance-coach skill in the profile, ships deterministic helper scripts, reads an athlete profile created by its init script, and requires the agent to state when data is unavailable instead of inventing it. It needs a separately installed Intervals.icu MCP server for training data and Python 3.10+ for its bundled scripts; with neither it works only from numbers you type in.

Bridges the official COROS MCP server into DeepSeek Harness with a self-contained local stdio-to-HTTP bridge and on-device OAuth login, exposing 34 COROS tools on the current endpoint for sleep, HRV, resting heart rate, recovery, training load, activities and FIT files, plus workout and training-plan CRUD and scheduling.

ModelScope connector: OpenAI-compatible chat models, a Models-page provider entry, and a composer magicube balance while a ModelScope model is selected.

A credit meter for XD Pool (WorkBuddy) accounts that shows each account's remaining credits and package breakdown, measures today/month/all-time consumption from balance snapshots, and warns about credits expiring within 3, 7 and 30 days. It reads its data from the XD Pool plugin (XDTrees/dsh-xdpool) and shows a data-source-unavailable state when that plugin is not installed.

Shows the RouterAI credit balance and the spend since the session started in the sidebar footer. Requires a RouterAI API key (`ROUTERAI_API_KEY`) and a model provider whose config key is `routerai`; without them it shows `--`.

Adds a 12-hour peak/off-peak billing strip and the account balance above the composer, with today, 7-day and 30-day spend per model.

Minimal DeepSeek balance badge in the DSH Web sidebar footer, with a peak/off-peak pricing line, low-balance warning and manual refresh. The API key stays in the host process and the balance route is limited to private-network clients by default.

Shows the DeepSeek account balance and unused bonus credit in the session header, refreshing every 20 seconds and whenever the window regains focus.

Durable workspace Plan/Backlog todos for DeepSeek Harness, with a sidebar board, agent tools, and /todo commands.

Narrow 40px icon rail for the DSH sidebar: panel icons move into a single column so the session list gets the full width, session titles gain a three-level hierarchy, and the DeepSeek account balance sits just above the sidebar footer with today and cumulative spend accounted from balance deltas, plus value-based colours. Follows the DSH 0.2.2-rc.2 sidebar structure (the two-column grid now lives on the nested scrollArea), so icons and the workspace list stay side by side.

Replaces the DeepSeek Harness home screen with a configurable palette and background image, a custom hero and sidebar logo, default 3×2 decks of recently used workspaces and quick links (local Chrome history or a custom list), and up to four API-usage widgets (DeepSeek balance, MiniMax, Cline Pass, OpenCode Go) refreshed on a configurable interval. The decks stay on the new-session screen and every option is edited from a built-in settings panel.

A named-agent-team chat workbench for DSH: per-member model routes with fallback and 3D mascot avatars; private and group chats with mention relay and bounded parallel lanes; a whole department as one agent whose lead dispatches sub-agents that run in parallel on their own models before the lead synthesizes; record search with hit highlighting and today/yesterday/earlier buckets; a four-layer memory system (raw chats → atomic facts → scene summary → user persona) recalled by hybrid BM25 + embedding RRF; plus an avatar library and local/cloud speech input. Hybrid recall needs an embedding endpoint (default: local Ollama qwen3-embedding:0.6b; otherwise BM25 only); speech input needs a local jianchang512/stt service or a cloud ASR key, otherwise it falls back to the browser speech API.

A simple, information-dense DeepSeek status display for the DSH sidebar: balance, usage, and peak/off-peak countdown. Requires the DSH Web GUI with a signed-in DeepSeek account.

Trims each tool result to a token budget derived from the active model's context window (ratio of window minus reserved output, clamped), keeping head and tail with a notice on how to fetch the rest; mount globally or inside one agent preset.

LAN pets for DeepSeek Harness. Neighbors' pets visit your desktop, and yours can visit theirs with a note or file along. Desktop pets need DSH Desktop (Electron) on macOS or Windows; on Linux or the web profile, discovery and visits still work but the desktop layer is unavailable.

Clears four kinds of idle-only system notices (approval policy, working directory, plan mode, plugin changes) that DeepSeek Harness counts as active work, so quitting with no running tasks no longer warns about interrupted work.

Audits and evolves agent context at four scopes: a single repository, a group of related repos, a knowledge hub, or the whole machine. Discovers projects, instruction files, skills, tools, knowledge, project state, evidence and credential references; reports drift, duplicate facts, broken pointers, paths left unverifiable in prose, budget overruns, unindexed knowledge, stale state and unregistered credentials; scores each scope on the dimensions that belong to it; and turns findings into reviewable migration plans that back up every file before moving it. The format contract behind the metrics is written down as a versioned spec (CTX rules, markdownlint-style IDs) that follows open conventions — AGENTS.md for instruction files, llms.txt-shaped link indexes. Also scaffolds a deliberately small hub for durable knowledge and project state. Ships a stdlib-only Python CLI plus two skills, with credentials referenced by path and never printed. Requires Python 3.9+ (stdlib only) and DSH 0.2.1-rc.2 or newer, below 0.3.0.

Wraps the separately installed agent-desktop CLI (`npm i -g agent-desktop`) as nine DSH tools (ad_status, ad_list_windows, ad_snapshot, ad_find, ad_click, ad_type, ad_press, ad_launch, ad_close_app) so an agent can list windows, read an accessibility tree, click a snapshot ref, type text, send keys, launch an app and close it gracefully. The plugin adds no automation logic of its own: it validates arguments, calls the CLI and returns the parsed JSON envelope, and asks the caller to take a fresh snapshot instead of retrying a stale ref, so a click never lands on a shifted target.

Treemap disk usage map with linked cleanup suggestions, a reviewed trash plan, and an append-only audit log on macOS.

Zhinao (second brain) for DeepSeek Harness: Markdown-authoritative layered memory (L1 long-term model gated by user confirmation / L2 knowledge with mandatory sources / L3 recent status under daily quotas), a local recall ledger explaining every injection, lifecycle engine with weekly review and a Web panel. First-run wizard initializes a blank brain, adopts an existing Zhinao directory read-only, or fills in only missing skeleton files; existing files are never touched.