Plugins

Deterministic tool-definition slimming for DeepSeek Harness: fewer repeated tokens per request, stabler prefix cache, plus a panel with provider-measured hit rate. · 压缩每次请求重复发送的工具定义，稳定前缀缓存，面板可调并可看实测命中率。

DSH client plugin for dsh-mesh: turns a signed, content-free wake ping from the hub into a real session turn, with a settings page, inbox card and history cache.

Encrypted, verifiable history keeper for DeepSeek Harness: snapshot, verify and restore your sessions. No git, no external binaries.

Notepad — a per-workspace floating notepad in the DeepSeek Harness session view with a dedicated settings tab (themes that also color the settings page, flat / frosted-glass surfaces, fonts, save mode, AI prompt presets with a read-only preset preview, bu

Quality-of-life enhancements for the DSH workspace sidebar: hide workspaces and their sessions, reveal them from the view options, and open any workspace folder in Finder, Explorer or your file manager. Display-only: workspace, session and archive data ar

Ten full-canvas animated backgrounds for the DSH conversation pane — starfield, code rain, circuit, sword-fight, cat, girl, fighter jet, racer, snow and bubbles — one random theme per turn, with a text-column veil that keeps the chat readable.

Read local EPUB, TXT and Markdown novels below selected native DSH chat regions on explicit expansion, preserving real reasoning and tool output without sending novel text to the model. Currently works only with DSH 0.2.1-rc.2.

Forward WeChat chat records into DSH sessions: a Windows share target captures the merged-forward ZIP, an in-GUI overlay picks the destination workspace and session, and the ZIP arrives as a file-attachment user message the agent unpacks and reads. Media entry extensions are normalized by magic bytes (WeChat ships PNG/HEIC bytes named .jpg), and failed deliveries preserve the batch for retry. Requires Windows 10/11 with Developer Mode on and the repo installer to register the share target.

Drive a dsh agent from Telegram: one chat topic per project directory, voice and photo input, ordered per-topic job queue, per-topic goals and task checklists, live progress message, command menu in en/ru/zh, and control commands for models, plugins, sessions and logs. Requires a Telegram bot token and your numeric user id in allowedUserIds.

Press once to refresh the interface and swap plugin code in place without restarting; running turns, queued messages and unsent drafts are kept.

DSH Desktop workbench for official MATLAB MCP experiments, verified PNG figures, MATLAB-computed metrics, and PID parameter sweep comparisons. Requires Windows, a MATLAB license and the separately installed official MATLAB MCP Server.

Pre-publish gate for DSH plugins. Turns every silent failure in the publish path into a deterministic check: package.json completeness (missing exports\["."], files omitting lib/ so the tarball is empty, no "type": "module", client declared without a ./client export), UTF-8 BOM, JSON/YAML parse errors, cordis.patch.yml insert rows written at the wrong indent (the bug that makes a plugin install and never mount), insert name mismatches, missing entry exports, require() inside ESM, missing LICENSE or README install line, missing tests, unscoped npm names that return a 403 no granular token can diagnose, versions already on the registry, and missing GitHub repos. Six agent tools; every finding carries its fix. Read-only, no LLM.

Read-only DSH profile doctor: the seven config-tree checks (config-clobber, dead-patch, tool-collision, plugin-not-mounted, plugin-stale, entry diff, explain) plus eight ecosystem checks learned from real failures — pnpm store version conflicts that break `dsh plugin add`, patch insert names node_modules cannot resolve, broken junctions, duplicate-capability plugins, orphan backup dirs, GitHub-source deps, npm publish-scope hints, and the minimumReleaseAgeExclude gap. Four agent tools; every finding carries its fix.

OpenFOAM case doctor for DeepSeek-Harness. Six agent tools: audit a case for 25 classes of silent failure (steadyState paired with timeStep writes, limiters set to 1 that let k/epsilon go negative, unbounded advection of transport scalars, nNonOrthogonalCorrectors 0 on a skewed mesh, loose solver tolerances, blockMesh vertex/block mismatch, stray time directories, purgeWrite 0); read any dictionary as JSON; survey a sweep of cases; run checkMesh/blockMesh/snappyHexMesh/foamRun inside WSL with log parsing; parse logs into the metrics that matter (aspect ratio, non-orthogonality, skewness, residual floor, fatal errors). Deterministic text analysis — no LLM — validated against a real 82k-cell street-canyon snappy case.

Audit how each turn's context budget is spent. Breaks down what a turn pays for AGENTS.md, the skill catalog, tool schemas and runtime context; flags repeated injection of unchanged content as pure waste; ranks skill-catalog bloat (with 60+ skills installed, how many tokens of descriptions does the model read every turn) and tool-schema inflation per plugin. Deterministic — no LLM, so a wrong answer is a rule bug with a name. Four tools including a cross-session sweep.

Audit what context compaction actually lost. Reads the raw session event log and checks session-constraint survival (the standing instructions the user gave — do-not-X, must-Y, only-in-Z — and whether the summary still carries them; arXiv 2608.11242 measured ~17% survival across compactors), summary hallucination (paths, versions, ports and identifiers the summary invents that were never in the history it replaced), ghost references ("see above" pointing at removed text), token accounting (a summary nearly as large as what it claims to free), and unaccounted prunes. Four agent tools including a cross-session sweep that ranks conversations by how much compaction hurt them. Deterministic — no LLM, so a wrong answer is a rule bug with a name.

Windows toast notifications for DeepSeek Harness: approve or reject a pending approval from the notification while DSH is in the background, with an in-app fallback; also reports turn completion and errors.

QQbot, QQ机器人 — watches chosen QQ groups over OneBot 11, batches messages by a count threshold, summarizes each batch with a model already configured in DSH, and sends the per-group report to chosen friends as a QQ private message. Messages can be filtered by keyword and by sender; ships a settings page, a local web console and a summary archive.

Sends desktop notifications and custom HTTP pushes when tasks finish or need your attention.

Shows a Windows 11 toast notification when a DeepSeek Harness session turn finishes, and stays quiet while a DSH window is focused.

Nudges the main agent of a long-running DSH session to report progress at a configurable interval, with exponential back-off, stuck detection, a per-session Progress tab, and a manual "nudge now" action.

Limbus Company desktop-lyrics style toasts for session events: approval requests, turn start and turn end appear as per-character animated text above every window on Windows — ripple/drop entry and exit, jitter or wave motion, per-event glow colors, random left/right screen-side placement, {tool}/{outcome}/{title} message templates, all editable live from a settings page in the DSH GUI; the transparent companion overlay is auto-launched by the plugin host and respawned when it disappears. Windows only; requires running npm install inside the package companion/ directory once to fetch Electron ^33 for the overlay.

Turns GitHub, GitLab and Gitee URLs in the transcript into hover cards with cached repository metadata, and opens them in the DSH sidebar browser on click; plain URLs and GenUI blocks are covered without patching GenUI itself.

A repository graph tab for dsh-better-sidebar, forked from @linxin666/dsh-client-ui-git-graph 0.4.8, which linxin666 develops in zhu1090093659/dsh-web. Both halves own the /git route prefix, so this plugin and the upstream one cannot be installed at the same time - pick one. It discovers the Git repositories up to three levels below the workspace directory (at most 200, skipping node_modules, dist, build, out, coverage, target, vendor and dot-directories), lists them as a tree with branch, head, dirty-file and conflict counts, and draws the selected repository commit graph as an SVG lane canvas with curved cross-lane edges, merge diamonds, ref badges, row zoom and a pinned commit card. The tab writes to the repository: the Initialize Git repository button runs git init, and the conventions picker writes a GIT_CONVENTIONS.md document into the repository (or a workspace-wide .dsh/GIT_CONVENTIONS.md that nested repositories inherit) from six frameworks - Conventional Commits, GitHub Flow, Git Flow, Trunk-based, Conventional Branch, GitLab Flow. The host injects that document into the system prompt before every model step and flags the last 20 commit subjects that break the rules the document itself states; Chinese, English and mixed subjects are all accepted, and an optional strict mode writes strict: true. The composer branch chip, interactive graph dialog and worktree manager of the upstream plugin are carried over unchanged.