Plugins

Reads the local NTQQ (QQ 9.9.x) message database on Windows and Linux and exports group and private chat history to txt, csv, json, md, html or sqlite. Ships sixteen model tools: qqnt_status and qqnt_doctor for environment and blocker checks, qqnt_extract_key, qqnt_index to build or inspect the sidecar index, qqnt_list_groups, qqnt_list_chats, qqnt_search, qqnt_stats for totals and monthly counts, qqnt_export_group, qqnt_export_chat and qqnt_export_batch (a chat can be named the way the user says it, and ambiguity refuses instead of guessing), qqnt_jobs for OCR, speech-to-text, chunking and model downloads, qqnt_media for a chat's downloaded images, qqnt_storage for disk usage and allow-listed cleanup, and qqnt_ask plus qqnt_recall for retrieval with citations. The local web UI on 127.0.0.1:46728 adds an incremental SQLite index, six views (chats, export, statistics, ask, settings, models), a task panel, a single-pane layout on narrow screens, sender redaction, media bundling, per-day splitting, export presets and AES-256 password ZIPs; OCR, speech and embedding components install into the plugin's own virtualenv on demand. Needs python3 with sqlcipher3 (the settings page can create a private virtualenv), and the database key is read from the running, logged-in QQ client.

Long-term memory kept as blocks: one event per block, written to a local append-only event log with a block view and derived indexes, so a full history stays recoverable while only the matching blocks reach the context. The model saves entries with memory_remember and pulls the relevant few with memory_recall, plus one automatic recall pass before each step. Data stays in ~/.beme; no external service and no API key.

Six-tier memory plugin that captures each conversation turn, injects a stable memory block as a user snapshot, and compacts the host's own history into a single placeholder without calling a model.

Automatic session titles for DSH: names a session from its first prompt plus the first assistant reply ("project｜task") instead of truncating the prompt. Replaces the built-in session-title provider through a bundle patch, adds an "AI rename" entry to the session row menu with a preview dialog, and a settings page for the provider/model, the naming rules and a diagnostic event log. 49 unit tests.

Recovers a session the host already refused for exceeding the context window: it re-checks the gateway message the built-in classifier misses, takes the real window out of that message, and folds the oldest span into one checkpoint without a model call (tool-result pruning, the host compaction engine when it can shrink the projection, or a mechanical digest fold), then retries the failed step; a context_rescue tool and /rescue, /rescue-status expose the same path by hand.

Adds an Auto-rename row to each Session sidebar menu: the model reads the Session log, weights recent user messages higher (exponential decay, 3-day time constant), and writes a title of at most 10 CJK characters into the Session title. Also opens that Session menu on right-click, and exposes the same operation as the session_auto_rename agent tool. Requires DSH 0.2.1-rc.5 or newer.

Adds per-model hide/show switches for custom (llm-pi-ai) providers to the DSH settings page. Hidden models disappear from the conversation model picker and are archived verbatim, so they can be restored at any time; disabling or uninstalling the plugin restores everything automatically. Built-in DeepSeek providers are not managed.

Runs the locally installed ZCode CLI as a `zcode` model route in DSH: all traffic is signed by the official binary, the plugin never sends its own request, forges client requests or sends credentials off the machine (the plan token is decrypted in memory only to hand back to the official CLI), no API key is needed, and both a paid one-shot and a free app-server channel are supported. Since v0.3.0 it ships an HTTP MCP bridge so tool calls run through the DSH approval surface. Backed by a contract document and a node --test gate.

Adds the models of your own GPU Hugging Face Space to the model selector, starts the Space when you send a message, shows its start progress, and pauses it when idle or at a time limit. Requires a Hugging Face account with a GPU Space serving an OpenAI-compatible API, a Hugging Face token stored in the DSH vault as HF_TOKEN, and a paid plan for GPU hardware; Hugging Face bills the GPU time.

Adds the models available in a signed-in WorkBuddy desktop app to DeepSeek Harness as a model provider, with the CN and international apps as separate groups. It is a rebuilt derivative of the npm package dsh-workbuddy-connect@0.7.9 (originally corrinehu/dsh-workbuddy-connect); it registers the same workbuddy / workbuddy-ai provider ids, so it cannot be installed alongside other WorkBuddy bridges.

Uses Z.ai / ZCode Start Plan quota inside DeepSeek Harness as a GLM model provider (GLM-5.3-Flash, GLM-5.3) with tool calling and reasoning levels; the settings page shows quota bars, gifted-plan claiming, refresh opportunities and a redacted request log. Requires the ZCode desktop app installed and signed in at least once (the plugin reads its local credential store) and an account with an active Z.ai / Start Plan entitlement.

Retry failed model requests automatically: intercepts every request error and replays the same step after a 2s / 5s / 10s / 20s backoff ladder, so gateway blips and rate limits no longer end the turn.

Shows the remaining quota of the current Our Free Model (OFM) channel beside the DSH composer, with a panel listing every signed-in channel; exhausted channels are flagged and switching models switches the channel shown. Requires dsh-our-free-model with at least one channel signed in.

Records DeepSeek Harness token usage into CC Switch, one row per model call, sharing the dedup ledger its own DSH importer uses so the two never double count. Also backfills existing session logs on first run.

Sets a CNY spending cap for each DeepSeek Harness turn (¥5 by default) and stops the run before the next model request once the cap is reached; the switch and amount sit at the right end of the composer dock.

GLM Coding Plan usage quota ring in the composer dock - 5-hour and weekly credit windows with level, auto-refresh and one-click retry. Requires a GLM Coding Plan subscription and key, plus your BigModel organization and project IDs (the team quota API returns empty without them).

Shows OpenCode Go subscription usage (5-hour, weekly and monthly) in the DSH sidebar footer, with a countdown to the 5-hour reset. Requires an OPENCODE_GO_API_KEY credential (Settings → Models, or a refs entry in ~/.dsh/.credentials.yaml); without it the card stays hidden.

A tiny always-on floating tree of the DSH workspaces, conversations and subagents that are running, waiting on you or on a schedule — live running state, who is waiting for your call, per-row model, scheduled-run countdowns, other local dsh instances with their ports, and local claude/codex CLI processes with their codex subagent threads.

Zero-friction quick ask for DSH: select text in chat, Markdown/code, rendered HTML reports, or the better-sidebar editor. The default answer starts automatically (English prose is translated, a term is explained, a question is answered); typing cancels it. Answers stream from a one-shot completion on the main window's model, inherit a read-only conversation snapshot, read the selected file as context (up to 20,000 characters by default), and never write to the main session. One-line install; host routes, client slots, document renderers and the iframe selection bridge are separable extension points.

Two-level model picker for the DeepSeek Harness composer: browse providers, then models, search across all providers or within one provider, select supported reasoning-effort levels, and navigate with arrow keys, Enter and Esc.

Adds a sparkle button left of the model selector in the composer toolbar: one click rewrites the draft in place through the host and briefly offers an undo that restores the original text. A Prompt Enhancement settings section picks the lane — the conversation model (the provider/model the next message would use, falling back to the configured default) or a keyless Pollinations channel that needs no login and no token budget but allows one reply per 30 seconds. During that cooldown a click completes through the conversation model instead of being refused, and the first click after the window reopens returns to the keyless channel on its own. The rewrite keeps the draft's language, and a result in a different language is rejected rather than pasted over the draft.

A permanent status card in the top-right corner of the DSH content area: one line for where this turn's time went (model, tools, waiting), plus collapsible to-dos with a progress bar, the session goal (edit/pause/resume/clear), submitted plans that open in the preview pane, the files this session changed, subagents with their token and duration figures, and this session's background jobs: running rows carry a two-step stop button, finished ones fold into a count line that stays collapsed until opened; it yields space to the content as the area narrows and folds into a round button at the limit.

Renders a dsh-view fence in an assistant reply as cards — text, tables, charts, Mermaid diagrams and sandboxed HTML widgets — with an editable model prompt, a render_view tool and a bundled skill.

DeepSeek Harness (DSH) plugin: Embeds upstream quota and usage progress rings from your local Magpie gateway (127.0.0.1:3425) right beside the composer context meter, with session-aware active model tracking and 5-hour limit reset tooltips.