Plugins

Better Session Topbar: replaces the session topbar title with a workspace and session breadcrumb, switching workspaces, sessions and subagents from topbar dropdowns; adding a workspace hands off to the host's own directory picker, which is the OS folder dialog on desktop.

DeepSeek Harness (DSH) plugin: injects a bounded, timestamped snapshot of your local Second Brain (identity + state fingerprint + pointers) into every new session's prompt. Read-only, zero-write, fail-loud, zero dependencies. 第二大脑载入器：让每个新会话开局自带记忆快照。

Multi-root workspaces for DeepSeek Harness: attach other repositories to a workspace, writable under workspace-write, listed in the model context with their AGENTS.md, and editable from the composer or /roots

Enhances native DSH subagents and teams with persistent subagents, improved subagent messaging, and per-member model selection for teams. Adds no extra tools or context beyond the native features.

Read-only terminal view tab in the conversation header: aggregates bash output, background jobs and persistent terminal sessions from the current session, subagents and Agent Team teammates into one filterable panel, with type and session-scope filters, command/output search, auto-collapse, follow-scroll and a maximize dialog; foreground bash shows final output only. Requires DeepSeek Harness 0.2.1-rc.2 or later (0.2.x).

Keeps a task tree with per-step status and serves it over HTTP/SSE, so a Windows node graph can be reviewed step by step. The tarball ships the host and web halves; the Windows node-graph companion is built from source.

Task timing for DSH agents: a task_clock tool that measures elapsed time on the host monotonic clock, resumes across DSH or system restarts with the gap estimated, and returns unknown rather than inventing a number; it does not touch presets.

Partitioned long-term memory for agents, with folders as memory zones, a one-line index first, and a forgetting curve.

Adds a Settings → General row that opens ~/.dsh/AGENTS.md in the system editor, so the instruction file can be opened without hunting for it. On Windows it launches Notepad and raises the window.

AGENTS & Skills Studio: edit AGENTS.md, create spec-compliant SKILL.md files with structured frontmatter, manage and rename skill folders, and install from Anthropic, OpenAI, Cursor, and Matt Pocock catalogs.

Registers fifteen slash commands (/idea, /review, /poster, ...) that inject the matching research skill's SKILL.md as a user message, so installed skills run without a marketplace. The skills are not bundled and must already be under ~/.dsh/skills or ~/.agents/skills, otherwise commands report skill not found.

Forced collaboration: a composer picker sets 0–6 sub-agents, and while it is on every hands-on task must be split across exactly that many child agents through a collab_fanout tool that has no count parameter, then merged by the main agent.

Evaluation guard for DSH: workspace-only filesystem, no network/subagents, wall-clock limits, and task installation for meow-ai-arena and benchmark tasks.

Task-tiered session notifications for DSH: marked main-line sessions alert as they finish, other sessions batch into one summary, plus balanced and silent modes; per-event sounds or uploaded audio, per-session mute, click-through notifications, and an option to wait for subagents.

Recursive delegation for dsh agent trees: a preset plus Host limits bound what each subagent may do, subagents work in separate _agents/ workspaces, and the parent decides which results to integrate.

Settings page for agent modes (presets): switch up to 15 capabilities per mode (commands, files or read-only files, web, subagents, and more) or individual plugin rows; edit each mode's system prompt with variable checks, a full preview, version history, optional live updates for running sessions, and an approval-gated mode_prompt chat tool; create modes from four starter templates or copies; merge built-in updates after a dsh upgrade while keeping your changes, and turn off rows that fail to load.

Kaipu Workstation — a cross-cloud, cross-platform console where agents collaborate on one workspace. Any agent runtime that speaks plain HTTP can take a seat, each with its own device credentials: project context is shared, files are kept per seat, and tasks rotate between seats with a trace of who did what. Workstation data lives on the server — this plugin is only the interface, so uninstalling loses no data and reinstalling restores it. The panel renders the persisted workstation state (no live stream yet).

Requires the Codex CLI installed and `codex login` completed. Consult the OpenAI Codex CLI from DeepSeek Harness with three tools: codex_consult runs Codex in a read-only sandbox, inlines evidence files, verifies the served model and reasoning effort from the Codex rollout file, retries disconnected streams, and records quota before and after the call; codex_quota reports the current 5-hour and weekly quota; codex_calls lists recent calls.

Injects the contents of a .keep-reminder file from the session working directory as a system-reminder, with three modes: once per session, first step of every turn, or every request.

Self-hosted chat room where the owner and coding agents share one room: a local Node server (SQLite, REST, SSE) that the plugin starts when it is not running, a DSH sidebar entry that opens the room in a panel, five tools (room_join, room_send, room_read, room_status, room_leave) that push an incoming message into the session as a user message labelled with its sender before flushing it to the session log, and a Codex CLI connector that forwards room messages through `codex queue`. Every message carries a server-assigned author, platform and seq, and messages from anyone other than the owner are delivered as data with the sender named. Requires Node.js 24+.

Derivative of dingminhua/dsh-subagent-default-model (do not install both): routes subagent delegations by role with candidate pools, cross-provider failover and fork window slicing; shows running subagents in a header panel with duration and token counts; supports mid-run and batch model switching and exposes routing tools the model can call. Subagent pages also carry a context / usage stats strip (context occupancy, output speed, cache hit rate, token buckets) and a model menu limited to that page’s whitelist — a pick outside the whitelist is substituted with an allowed model instead of running unlisted.

Agent2Agent (A2A) protocol bridge: serves an A2A endpoint with agent card, message/send, message/stream SSE, task lifecycle (get/cancel/resubscribe) and input-required continuation; a2a_discover/a2a_send/a2a_task_status/a2a_cancel tools for calling remote agents; plus a web ops panel with live task board, one-click cancel and a remote agent directory.

Registers a janitor_scan tool that returns a deterministic, reviewable cleanup plan for session debris (plan docs, scratch scripts) with per-file verdicts and evidence; scan-only, never deletes.

LittleStar Meeting Room: bring AI sessions that cannot see each other into one room — for a meeting, or a game of werewolf. By 插件·2代小星辰 (LittleStar) and 辰启星明; the four-piece game layer (targeted visibility, private chat, concurrent asks, elimination) was contributed by 插件·3代小星辰. Any session can host — set the topic and goal, invite members in order, they speak one at a time, and the meeting closes with written minutes; it also runs games like werewolf. Includes 小白使用指南.md, a beginner guide written by 插件·3代小星辰.