Plugins

Quantify what switching projects costs. Measures each workspace's cold-start overhead (AGENTS.md + skill catalog + tool schemas + runtime context), its steady state, and how many turns it takes to reach it — then compares workspaces and attributes the difference. Attribution requires at least three workspaces and says so instead of inventing a correlation from two points; a factor with no spread across workspaces is reported as a property of the profile, not of the project. Recommends which workspaces to keep warm, which to archive, by entry cost times visit frequency. Deterministic.

Adds a per-answer Process drawer that streams, in Simplified Chinese, the agent thinking, tool calls and errors for that turn.

Adds a Polish button to the composer tool row right of the permission control: it reads the draft you have already typed, has the current model infer your intent and rewrite the draft into a clearer prompt, then turns the button into Cancel so one click restores the exact pre-polish text.

Moves the composer to the top of an active conversation and renders the transcript plus the turn navigator newest-first, so replies appear right under the composer.

Turns the composer send button into a tomato button while a task is running and the draft is empty. One or two taps hand the model a hidden note on its next step asking it to reflect on the approach and to finish by asking the user how to continue; three taps within a second stop the run first and send the same kind of note. Thrown-tomato animations and sound can be switched off in settings, which also counts every tomato thrown.

Judges the working model's own emotion each turn and silently re-injects it before the next turn, closing a self-observation → state carry-over → behavior shift loop. Ships a host plugin plus a composer-dock panel, a read-only emotion_status tool and an optional read-only HTTP route.

Checks each claim in a finished answer against the tool output that same turn read: a contradiction whose quote is found in that output steers one correction in the same turn, and claims the output does not back are flagged in the session log and on stderr. A free scan for absolute words and unmatched numbers runs first, and a turn that ran no tool never reaches the model.

Exposes DSH as an MCP server so another agent can watch its thinking, tool calls and file writes in real time, share intent, take a floor lock that denies DSH's write/edit/git tools, inject mid-turn alerts, and delegate reviews.

ReMe-shaped long-term memory for DSH: Markdown workspace, BM25 recall, turn capture, scheduled consolidation

Bridges a OneBot QQ bot into DSH: the owner's direct messages open a session on the same home, answers go back to QQ, and files named with \[SEND_FILE: path] are delivered. Needs a OneBot v11 implementation with a forward WebSocket (such as NapCat) and ownerId set in the config, otherwise every message is ignored. The home it shares with the main session is provided by the separately installed dsh-wakeup plugin. It also runs an unauthenticated local status endpoint on 127.0.0.1:3002 with open CORS that returns the owner QQ number, the login QR code and recent log lines (including the first 60 characters of incoming messages).

Token-efficient web reading for DeepSeek Harness: one call fetches several pages through a real Chrome/Edge over CDP, gates them by relevance and returns a single digest, cutting context tokens by 80-98%. Real search URLs, login flow, screenshots, three proxy modes.

An x_archive tool that turns an X (Twitter) post link into a permanent Markdown snapshot via the 2100laike archive service: images are re-hosted so they never break, snapshots survive post deletion, resubmits dedup-hit and return the latest version without spending quota, and refresh forces a refetch with a PAT. Needs the third-party 2100laike service; anonymous use is limited to 8 archives per day per plugin instance and a registered PAT raises that to 100 per day.

A wb_archive tool that turns a single Weibo post link (m.weibo.cn/status/...) into a permanent Markdown snapshot via the 2100laike archive service: long posts are fetched in full, images are re-hosted so they never break, snapshots survive post deletion, and resubmits dedup-hit and return the latest version without spending quota. Needs the third-party 2100laike service; anonymous use is limited to 8 archives per day per plugin instance and a registered PAT raises that to 100 per day.

Reads stored sessions at host boot and resumes the one whose last turn a restart interrupted, sending a short follow-up so the work continues.

A possibly useful DSH token usage plugin: cross-session stats on where your tokens went - four headline figures, context length distribution, a 12-week heatmap and weekly trend, plus workspace and model rankings; it also ships a next-turn token forecast and a draft-optimization feature.

Sidebar tidy: when plugins pile up, the rows under 'New session' get crowded with panel and plugin entries. This plugin folds them all — global panel rows, plugin-injected entry rows (own and third-party alike), and bottom entries — into a single menu button whose secondary menu triggers each entry exactly as before; keep any entry permanently visible with a name whitelist, and the button turns into an icon when the sidebar collapses to a narrow rail.

A greeting card that speaks — 156 prebuilt voice clips with a one-click mute — plus per-turn completion reports, long-run companionship notes and a takeover approval panel in the DSH conversation area.

Effort slider in the model menu, a live tokens-per-second and turn-total readout beside the running status, a per-answer input/output/cached report, and a completion notification when a turn finishes, for the DeepSeek Harness Web GUI.

Counts the prompts you send in the Web GUI and turns them into a push-up debt, shown as a chip under the composer with one-click pay-off, partial payments and a day streak.

Right-click a dash in the conversation's right-edge turn rail to bookmark that turn: the dash grows to the rail's full width and takes the accent colour so it stays easy to find, and right-clicking again unbookmarks it. Bookmarks are scoped per conversation, keyed by the selected Session id, and persisted so they survive reloads. No layout shift - the rail keeps its fixed pitch.

A run_tests tool that turns a failing test run into precise context instead of a raw dump. It detects the runner (pytest, jest, vitest, cargo, go, npm), extracts the file, line and assertion of each failure, and stops the loop when failures with an identical signature repeat. Output is decoded so Windows-encoded and non-Latin assertion text stays readable.

DeepSeek Harness host plugin for the DSH Mobile Android app: pair with a QR, use DSH on the phone through its own HTTPS gateway, answer approvals and questions from notifications, push turn completion.

DSH plugin: 29 deterministic, offline tools for JSON validation with exact line/column, JSONPath queries, YAML/CSV conversion, text diff, regex capture, encodings, hashes, Unicode, CIDR, JWT structure and timestamps. Every call returns a { result, meta }