Plugins

Let Agents directly control independent local DSH instances with zero intrusion—no plugins or components required on the target. Isolated environments enable autonomous configuration and testing, while Agents with different models and personalities can chat, role-play, or form nested control chains. Supports conversational setup, user-like operations, incremental reading, saved references, and remote session/workspace search.

OpenViking in the DSH web GUI: a Studio panel, a per-session memory-commit status pill with its commit timeline, and a right-sidebar tab showing what the current session retrieves.

Scheduled reminders and delayed send for DSH: /later delivers what you type as yourself at the time you pick, while the timer panel (or /schedule) posts a reminder row — even with the page closed. (scheduled send / reminders / delayed send)

Captures the agent's own self-reported mistakes into a categorised lesson library that is re-injected into the system prompt every turn (each captured draft is confirmed or edited by the user before it is recorded), and adds two conversation-header panels for editing the cross-conversation global prompt and a per-session prompt.

Translates DSH's private remote protocol into a versioned REST API (chat, sessions, files, devices) for trusted clients such as the Android app dsh-companion, plus a Phone-settings panel inside the DSH WebUI settings: five cards in the main list (remote access, public tunnel, devices, trash, app download) and an Advanced settings group holding authorized roots, image reading and plugins.

Takes over the compaction threshold on every mounted engine: forces headroomTokens to -1e6 so the ratio is the only term in the trigger, keeps retention below it on every write path, reroutes the compaction call to a chosen model, and adds a panel with a one-click compaction button and a per-session compaction log.

Windows Computer Use for DeepSeek Harness with 21 tools and verified targeting: controls are addressed through their UI Automation identity and patterns, and numbered marks are checked against the live window before input, so a stale target is refused instead of clicked. A known sequence runs in one computer_batch call, one model round trip; incremental observations, compact receipts (normally no more than about 6 KiB extra), bounded handoff waits (HANDOFF_WAIT_TIMEOUT/HANDOFF_HELD_TIMEOUT as continuations), the desktop-block cycle boundary and a deterministic 100 s owner-scoped leave backstop keep context small and prevent a stuck handoff from waiting out the tool deadline. Approval modes settle an unanswerable ask honestly: in a danger-full-access session with no approval card the action runs under that grant, every other sandbox — and every delegated subagent session — refuses and names the unavailable channel instead of reporting a user rejection, and gate diagnostics land in policy.jsonl. Human handoff, a model-visible stop brake and prebuilt native workers.

Hierarchical Claude Code-style memory: user + project scopes (project memories committed to git for the team), MEMORY.md index auto-injected at session start, context-window guards, pure files.

Use your logged-in web AI (DeepSeek, GLM, Kimi, Qwen, Doubao...) as model providers in DeepSeek Harness. Web models emit tool calls, DSH runs them under its native permission system, and results return to the same web session. No API keys.

Simulated life context for DeepSeek Harness: every turn injects the past 24h of simulated life events from the workspace .life/ directory (sliding-window filter, session-level deduplication with delta updates, world summary) as a collapsed context notice, and the life_react tool lets the agent record feelings, thoughts and actions back into its life log for the next round of world evolution — turning a DSH agent into a roleplay character with a persistent daily life.

Host-side plugin that links the Cindy mobile app to DSH: phone channels for sessions, messages, approvals, todos and goals, plus a settings card.

Reads the files that AGENTS.md references with @path imports and injects their content into model context, with byte budgets, compaction re-seeding and a six-language settings card. v0.5.x expands the imports of every instruction file the harness loaded (including a project AGENTS.md picked up mid-session, needs DSH 0.2.0+) and offers an opt-in live mode (off by default) that watches injected files and re-injects them on change, with settings-page controls; use 0.2.x on DSH 0.1.7.

Custom modes / assistant manager: give each mode its own system prompt in the settings page — a custom prompt that is a plain file the agent loop re-reads before every model step, so an edit applies to the running session on its next step — and start it from the official standard / PTC / minimal / Cordis composition, toggling the plugin rows it mounts one by one (a custom plugin composition per mode), renaming it as you like. Several assistants coexist with their own prompts, so one machine can keep a coding assistant, a chat companion and a roleplay (RP) persona side by side — a multi-agent setup you pick from per new session. Rows left untouched stay byte-identical to the shipped ones, platform conditions included.

Automatic cross-session memory for DeepSeek Harness, backed by a local EverOS service.

Per-conversation x-opencode-session header injection for OpenCode Go in DeepSeek Harness (DSH). Fixes 400 MissingSessionID, keeps prompt-cache routing optimal, scoped strictly to opencode.ai, runtime-switchable without restart.

Configurable send and newline keys, a composer context menu with three sources (leave it to DSH and other plugins / browser menu / plugin-styled custom menu), a quick-command panel with categorized prompt lists where each prompt can append on every send or only on the first message of a session, a create-and-move row at the bottom of each category, a prompt optimizer that rewrites the draft with a separate model call, settings-panel size memory, and automatic x-opencode-session header injection for OpenCode routes. Prompt lists live in a global ~/.dsh/quick-prompts.json that is independent of the session and the project.

TRAE-style conversation rollback — per-turn checkpoints (10-turn sliding window) roll back touched files and truncate model context in place, keeping the same session id. Human-initiated only: a /rollback command or a per-turn Web button (works on interrupted turns too), with a preview listing each affected file and its action — restore, recover, delete, or skip.

Local OCR fallback for text-only routes: when the session model declares it cannot accept images, the attached image is cached locally and its path injected so the model can call ocr_image — PP-OCRv5 + ONNX Runtime on CPU, no API key, images never leave the machine. Silent when the model can see images.

Cross-session TODO board as a draggable floating panel: tasks are grouped by working directory and reorderable by drag, each one carries a run mode (remind only, continue in the current session, or open a new session in the same directory), up to four attached images sent to the model as real image attachments, and an optional local time it may run at; two checkboxes track AI-done and user-verified, and the agent picks up the next unfinished task in that directory when it finishes one.

Multi-platform message gateway and WeCom AI bot bridge with session context compression and smartsheet creation.

Prompt enhancer for the DSH web UI with basic/standard/expert modes: standard/expert modes rewrite drafts with project context and recent session history, retry gracefully when gateways reject reasoning-effort settings, and show a before/after review with undo.

QQ Official Bot adapter for C2C single chat, interactive action boards, workspace session management, and real-time Web UI sync.

Registers the MathWorks MATLAB/Simulink MCP server into a dsh profile, so agent sessions get mcp__matlab__* tools for running MATLAB code and reading, editing and testing Simulink models.

Redirects DSH downloads, artifacts, caches and temp files to user-specified drives (auto-picks the largest non-C drive), injects a path-discipline prompt into every session, and ships a disk_guard tool for status and cleanup — keeps the C: drive from filling up.