플러그인

Adds the real CNY cost of each assistant turn, plus today's running total, after the timestamp in DeepSeek Harness's completed-turn footer. Amounts are folded from the session log's usage with a local price table - no billing API, no network call, no extra tokens.

Turns scattered thinking notes into a growing constellation of themes, claims, relations and open questions, plus a written synthesis of the argument.

Per-turn next-step suggestions under each reply: a 2x2 grid of up to four clickable actions, projected from that turn's own tool result instead of a shared per-session store.

An animated cat companion for the DSH chat UI. It reacts to session states, shows editable click-to-read lines and per-turn token counts, and supports dragging with a squish effect, left/right flipping and Ctrl+wheel resizing.

Article Review page in DSH's native right sidebar: turns Markdown, Word, PowerPoint and PDF articles into sentence-addressable text, lets you comment on any sentence, and hands the open review to the agent as a composer draft. Six agent tools let the model read every comment, revise the article (in place for Markdown, as a sibling <name>.revised.md for binary formats), resolve the comments, and record recurring wording preferences into a long-term writing-style memory injected into later prompts. Extraction includes a dependency-free DOCX/PPTX reader; PDF text extraction runs locally through the host's Python + pypdf (or poppler's pdftotext) and is unavailable without one of them. No network calls, no telemetry.

Turns every .pluginlet file under a repo's .dsh/pluginlets/ (directories configurable) into its own sandboxed Conversation View tab, gated by a capability grant bound to the file content hash.

DeepSeek-Harness 插件：轮次收尾时若模型未声明目标状态，注入续行提醒而非放行。处理思考泄露、工具调用畸形、上游断流导致的任务静默中断；上游中断类型可重试错误耗尽后自动续行，默认上限 3 次，跨进程共享计数。含设置界面与判定留痕

Copilot-style file review for DeepSeek Harness: inline per-edit diffs, whole-file Changes sidebar, shell write guard, Keep/Undo and persistent per-turn Restore Checkpoint.

Sayaka Miki desktop pet for the DeepSeek Harness Web UI: twelve hand-drawn sprites follow the live session state and swap instantly, with a tired slump after a turn, a furious second stage after ten pokes, and idle sword-polishing; repeated states change her line and a voice line plays every five to eight encounters.

Feishu (Lark) bot for DSH remote control — messages become session turns, results return to the chat, and ask-mode approval requests can be approved or denied in Feishu.

The mobile client for DeepSeek Harness: use the DSH running on your computer from your phone. The Android app signs in with the same account, picks this computer from the device list and opens its DeepSeek Harness UI on mobile. The desktop plugin handles registration and the connection, and nothing on the computer needs an inbound port. The UI adapts on its own: on a phone it becomes a touch-friendly layout, and returns to the desktop layout when a foldable opens or the window gets bigger. A PC-side panel shows login state, device quota, device list and the app download QR code.

A Feishu/Lark channel for DeepSeek Harness: WebSocket inbound messaging that steers a running turn instead of queueing behind it, Markdown tables rendered as native Feishu table components, code blocks kept on their own lines, 13 Feishu tools, voice input and output, media upload, a model fallback chain, and a web settings page split into six grouped tabs. Runs on a self-built Feishu app over the long connection, so no public endpoint is needed.

Security review for DSH plugins: reads the manifest, install hooks, native dependencies and source for network egress, dangerous APIs and credential paths, cross-checks the curated market registry, and returns a 100-point verdict with itemised findings. Read-only — it never executes the audited code.

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.

Blocks the end of a turn while temporary files from that turn are still lying around: it registers a listener on the serial agent/turn-stopping hook and, if anything is left, steers the model with the file list and two ways out — delete them, or justify each one and move it somewhere deliberate. Candidates come from two narrow temp-file sources, plus a backup-retention channel that is on by default and flags excess .bak copies of files touched this turn. The first is the .agent-tmp/ convention directory under the session cwd: whatever is left in it is reported outright, because that directory exists only for throwaway files. The second is the paths this turn actually wrote, captured through the fs/write-intent hook; those are reported only when they sit under that cwd, fall outside .git, node_modules, .tmp-sessions and out, match a temp-file pattern, and the file is still present — which is what keeps junk already sitting in the workspace from being resurrected as a complaint about this turn. One steer per turn, subagent sessions skipped, nothing deleted by the plugin in the default steer mode (only the opt-in mode: prune deletes backups), and the check is readdir plus stat only.

Adds a bell to the sidebar header: yellow when a conversation waits for approval, red when a turn was interrupted, blue when a conversation completed; clicking it opens a status summary panel and a session board, and a status with exactly one session opens that session directly.

Turns your local WeChat group chats into a followable todo board: a three-column kanban (todo / doing / needs you) with LLM summaries, automatic flagging of messages that @-mention or reply to you, drag-to-change status, and local-only storage. Needs a local Python 3 with sqlcipher3 and zstandard, plus the existing rion-wechat-reader config.json and keys.json for the WeChat databases; the plugin does not extract the keys itself, and summaries send the quoted chat text to the model you choose.

Tells the parent agent when a subagent died from an upstream failure instead of finishing: it listens to agent/error on every agent, waits a short settle delay, and if the child is still not running delivers a followup turn to the parent carrying the child id, the failure and the options the parent has, with a cooldown and a per-child cap so a flapping provider cannot flood the parent.

Right-sidebar flow editor for prompts, model switches, commands, condition branches and loops, executed only at turn boundaries so an in-flight request keeps its model.

A fact layer for daily heartbeat turns: diffs the curated plugin market for newly published entries and the Chinese education SRC calendars (EDUSRC rounds, Beijing education SRC seasons, the standing vulnerability-hunting exercise) for new items. Reports deltas only — the first call seeds a baseline and stays quiet.

Synthesises speech through Alibaba Bailian's WebSocket TTS and returns a WAV file plus a preview URL, with an optional per-turn hint that nudges the model towards voice. Needs a Bailian API key and ffmpeg, and the text is sent to Alibaba Cloud; the plugin does not send messages itself — delivery is left to the session's own transport.

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.

Model failover router that queues every connected DSH model into priority templates, retries a failing candidate before switching, cools down rate limits briefly, quarantines failures per session, and auto-sends a continue message when a turn fails or hits the token ceiling.

Slims PTC SDK tool definitions to plugin-level summaries and tool names, then returns the full visible tool pack when the model calls unwrap with an exact plugin ID, without changing scope or permissions.