Plugins

Read-only CNKI bibliographic search for DSH: four MCP tools (search, detail, export, status) drive your own Edge/Chrome through a native-messaging bridge, with no PDF downloads, a 3-second minimum interval and a 300-action daily cap. Requires pip install wensu plus wensu setup and loading the browser extension before the four tools appear.

Park whole tool packs out of the session's tool surface and mount them back with one command. Cuts the tool table from 99,789 to 48,144 characters (51.8%) without uninstalling plugins or restarting. Reversible per group, per agent.

Adds a Chat mode beside Work with a user-defined persona, long-term memory extracted after each turn and recalled by topic, and stickers it can save and send.

Port of the V2 memory pipeline from OpenAI Codex to DeepSeek Harness: a background agent extracts each idle session, consolidation folds the extractions into one memory_summary.md behind a git baseline diff, new sessions get that summary injected, and cited memory flows back as usage counts. A Settings page chooses the model, each stage's reasoning effort, and the idle, age, concurrency and timeout limits.

Nothing stays in the system prompt: a hit returns titles only, a body loads only when read, and project rules can grow themselves.

Three-layer memory for DeepSeek Harness — a small always-on core block plus on-demand topic files (profile, preferences, decisions, lessons), with source-annotated writes and supersede-not-overwrite history, so preferences and project constraints stop drifting between sessions. Local-first: no vector DB, no server, no outbound calls.

Memory that outlives the session, kept as plain daily log files: one markdown file per day plus a generated index the agent greps instead of re-reading. Registers the annals skill, injects the write-it-down rules into the system prompt, and exposes three tools — annals_init builds the log directory, the index generator and a separate Git repository in one call; annals_query full-text searches past entries and returns them with file and line anchors; annals_links resolves which entry cites which. Reminds the agent once per turn when it changed files but wrote no log that day, and once more when the log was updated but the index was not. Adds a settings page for the six switches.

Turns corrections into permanent rules: explicit remember: triggers and auto-detected corrections pass a confirm bar (save globally / per project / edit / skip) and are written into a marked AGENTS.md section. New rules take effect in the current session and load natively in future ones; writes are deduplicated and version-guarded, and secret-looking content is refused.

Long-term memory on top of mem0 and a local Chroma store: user profile and recalled facts injected through the system prompt, capture at session end, a nightly pass that dedupes, merges paraphrases and resolves contradictions, a persistent topic graph written as Markdown cards, periodic export to a git repository, and a full-page memory UI with search, bulk merge, snapshots and undo. Needs a manual Python venv setup plus LLM and embedding API keys before use.

Progress tab for the DSH right sidebar: the turn status of the session and the progress long-running tasks reported, in one tab beside the conversation.

Flat list of recent sessions across every workspace, with last-question previews, time buckets, full-text search, and notes that resurface when you reopen a conversation. Fork of huangxp12/dsh-recent-sessions with a native-style UI rewrite; both register the same recent-sessions panel id, so do not install both.

Deletes a whole DSH session or one completed turn from the session log: confirmation dialog, live-agent teardown, path and identity checks, pre-rewrite backup, and later turns renumbered.

Lists the attachments the current session actually received through the WeChat bridge and adopts one into the workspace on demand.

Adds a Timestamp toggle to the composer; while it is on, each message you submit is followed by a hidden context message carrying the current local time, so the model knows when you sent it. The switch is remembered per conversation.

Per-conversation system prompt, turn-based context folding, sampling overrides and a low-token pure chat mode for DeepSeek Harness.

Collects the sessions that need you into one dock at the bottom of the web GUI sidebar: running sessions, sessions waiting on an approval, plan review or answer, and finished runs whose result has not been opened yet. Rows are sorted by last update and clicking one switches the main view to that session; it reuses the shipped useSessions and useSessionStatus hooks rather than adding a data source, polls nothing, stores nothing and modifies no session. The header dot pulses while anything is running or waiting and settles to a static dot when every row is only a finished-but-unread reminder; with no such sessions the dock collapses to one quiet line, and a collapsed 56px sidebar shows a 36px badge with the count instead of the list. It registers in the sidebar footer slot, so it is part of the layout and does not overlay the conversation list, leaving the existing footer entries below it.

Persona plugin: a slowly moving 0-100 emotion meter, its own worldview, a per-turn diary, a relationship network, proactive messages, and an inner-voice line beside the cost row.

Offline LAN chat inside DSH: direct chats, group chats and file transfer between machines running the harness on the same network, with UDP auto-discovery, CIDR scanning to add contacts, an optional access code (unset by default), local-only contact aliases, per-message and per-conversation deletion, and a configurable per-file limit. The host opens its own HTTP/WebSocket port for peers, and the plugin imports only node built-ins.

Adds a connection-test panel under Settings -> Models: reads the deployment's model catalog and probes each configured model with one minimal chat round-trip (a short "ping" with a small output cap) through the Host LLM service, verifying connectivity with a real request. Each row reports success/failure with the harness failure code (NO_ADAPTER, MISSING_CREDENTIAL, AUTH, RATE_LIMIT, TIMEOUT, ...) plus first-token/total latency; single-model and concurrent test-all (3) are supported. Credentials are resolved inside the Host process and never reach the browser; probes are independent per model behind a same-origin-fenced route with a body-size cap. Pure JS, no build step, MIT.

Rates how hard the current message is and shows the cheapest model and reasoning effort that still fits, above the composer, before you send.

Bailian (DashScope / Model Studio) model catalog in the DSH settings page: lists every model the configured API key can reach, groups them by model family, probes whether a model is actually callable (an unopened model answers 403, which DSH reports as an invalid key), refuses to add models that fail the probe, and shows the active session's context window and remaining budget. An sk- key has no balance endpoint, so quota here means context quota.

Stardew Valley theme (full-skin edition): replaces the served web dist with a self-contained pixel-farm frontend — Junimo dialog, farm-select agent presets, quest journal HUD, session-as-day valley calendar with seasons and weather particles, chiptune SFX and desktop pets. npm: dsh-theme-stardew-fullskin.

Orihime golden-hour skin for DeepSeek Harness: a two-layer illustrated backdrop with a sidebar workshop for composition and sharpness, amber message bubbles and a cream letter-paper composer.

Stripe Style theme for the DeepSeek Harness Web UI: a 40px purple grid, layered soft shadows, 12px cards, an 8px control scale, and an in-settings theme row with Daylight Violet and Midnight Navy palettes plus one-click factory restore.