Plugins

Cross-session long-term memory for the DeepSeek Harness: a two-phase extraction/consolidation pipeline over a file-based memory workspace, recall injection, a citation feedback loop, lease-based job semantics, plus a settings page, a sidebar status panel

Keeps an unfinished task list visible across turns. DSH clears the todos projection at the start of every turn and expects the model to rewrite the whole list, so when a turn is interrupted the plan is durably in the session log but permanently gone from the UI — and todo_write is write-only, so neither the user nor the model can read it back. This plugin registers one ordinary session/event listener, remembers the last list per session, and when a turn/start arrives whose remembered list still has unfinished work, re-appends that list so the projection stops reporting an empty plan. No fork of tool-todo and no waterfall: a plugin cannot replace the projection (the registry throws on a stateVersion mismatch), so appending the event the fold reacts to is the reachable seam. Deliberately narrow — only unfinished lists, at most one event per turn start, fails open. Verified by a controlled A/B on real sessions: identical runs differing only in whether the plugin was installed, folded through the real projection definition. Zero dependencies.

Injects your own saved context entries as a Remember row after each user message, per turn or once per session, with a Settings editor and a manager page.

Local token usage and cost dashboard for DeepSeek Harness: four-segment token breakdown (miss/cache-hit/cache-write/output), cache savings estimate against the miss/hit price gap, per-session drill-down with context-growth curves and anomaly flags, global filters with shareable URLs, monthly budget progress with month-end projection, and day/month aggregate fast-path queries (O(days)) with incremental rescan.

Screenshot-review skill for DeepSeek Harness: the agent opens a frontend page, screenshots it, views the screenshot itself, revises the code, then re-screenshots — iterating until the page visually matches the user's requirements. Ships as a dsh.bundle, so it installs with `dsh plugin add`.

Context curation for DeepSeek Harness: content-aware tool-output slimming, dead-log sweep, pinned user requirements and in-flight work, archive-then-compact with an audit trail.

DSH Web 任务面板插件：把需求、队列、子 agent 执行、自动复核、验收返工和定时执行做成一个可视化闭环。

Local semantic index over workspace documents (MD/PDF/DOCX/TXT): FTS5 BM25 + local embedding hybrid retrieval, hit citations with line numbers, and incremental updates.

DSH 智能上下文压缩插件 —— 在有限窗口内管理认知资源，让 Agent 记住真正重要的信息。 Smart context compression for DSH — managing cognitive resources within limited windows, so your Agent remembers what truly matters.

Global experience archive for DeepSeek Harness on Windows: probes local tools (with versions and PATH-first hits), convenient paths, network latency and disk speed; folds per-session stats from session/event; LLM-extracts durable facts; env_profile tool with remember; byte-stable summary injected every turn to speed up sessions and cut token use.

Cross-session event-cue memory for DeepSeek Harness: auto-extracts task/conversation cues and user language habits from every turn into a persistent global cue bank, then re-injects remembered context on topic-switch detection (keyword or vector matching).

Incremental diff reviewer for DSH: after approving a batch, only new changes since the checkpoint are shown, with a Web review panel and `/review` command that inject feedback back to the agent.

Long-term associative memory for DeepSeek Harness: a local plain-text vault of raw episodes, typed memory cards and entity pages, configured from 设置 → 记忆图谱, and a Memory tab inside a conversation that draws what that session remembered. No build step; the

File-state cache skill for coding agents: one explore call replaces ls+grep+read, reads remember they already happened, and edits return a diff plus the exact changed line range instead of a full re-read. State is a JSON cache on disk, so it survives across tool calls and restarts. Requires Python 3.8+ (standard library only).

Four-layer memory for DeepSeek Harness, ported from the memory design of NoMindSama/qq-dsh-bridge (MIT) with the QQ channel removed: an append-only dialogue log, a weighted long-term fact store with decay and one-way archiving, self-notes the agent writes about itself, Chinese bigram recall over the raw log, and incremental AES-256-GCM sync to a private git repo where only ciphertext is stored. Archived sessions drop out of recall without losing their history.

Draws every question and answer in the DSH Web conversation blurred until the pointer rests on that row, and locks the mask to always-blurred while the window is unfocused; one control below the composer switches between off, hover to reveal, and always blurred, remembered per browser.

A desktop companion for DeepSeek Harness (dsh): she lives on your real desktop (on top of your windows or inside the wallpaper), talks with you by voice and remembers you. · DSH 桌面女友：住在你真正的桌面上，能语音聊天、记得你。

Agent-team task workbench for DeepSeek Harness: a sidebar board and requirement view track tasks across backlog, execution, review and done, dispatch work to coding agents, and run multi-role team orchestration on a local SQLite store.

Measures one university department's research output: matches author affiliation strings against a college's known name variants (detecting same-named departments at other universities), backfills papers that only name the university by faculty author IDs, classifies each paper by journal tier (CAS partition with Top flag, JCR quartile, CCF, Chinese core via a CNKI export merger), splits staff from student first authors, and writes a workbook with a dashboard sheet, a filterable paper list, a teacher-by-direction matrix and per-teacher statistics. Student authorship is inferred from affiliation and staff lists, not verified against enrolment records. Requires Python 3.9+ with pandas and openpyxl (pip install -r requirements.txt); papers come from the OpenAlex API, whose free per-IP quota is limited, and Chinese-journal coverage needs a manual CNKI export.

Floating TTS speaker in each conversation corner: when an AI reply finishes, edge-tts reads the summary aloud in the background with no popup windows; the toggle is remembered per conversation. Requires Windows, Python 3.9+ with the edge-tts package installed, and network access to Microsoft's Edge TTS service; summary is the default mode.

IMAP mail triage and cleanup for DeepSeek Harness: incremental read-only scan, a hard-protection rule layer plus optional local-model classification, a Markdown report and a hashed manifest, an explicit approval card, and only then a move to the Trash mailbox — no expunge.

Indexes public AI-company job boards (Greenhouse, Ashby, Lever, Workable, SmartRecruiters), scores postings for AI relevance, extracts the requirements from each job description, and produces a scheduled daily Markdown digest.

Three photograph skins for the DSH Web GUI — River Night (三江夜), Lingyun Dusk (凌云晚照) and Emei Mist (峨眉夜雾). Each paints its Leshan photograph across the window and the conversation column under an appearance-keyed scrim; one engine derives the other 119 --dsw-* alias tokens from six colour inputs per appearance, Alt+K switches skins, and the choice is remembered in localStorage.

A persona that grows from experience: remembers what happened, becomes someone from it.