dsh-undo-savepoint
DSH undo/rollback system: snapshot config files on change, undo/redo the last action from the WebUI or by chat, and roll back broken plugin trees without reinstalling. Works even when DSH fails to boot (offline CLI + GUI).
133 results
DSH undo/rollback system: snapshot config files on change, undo/redo the last action from the WebUI or by chat, and roll back broken plugin trees without reinstalling. Works even when DSH fails to boot (offline CLI + GUI).
Turn-level conversation and workspace rewind for DeepSeek Harness, powered by a persistent Change Ledger
DSH 插件:真正便捷无感的同窗口内对话回退,从不新建分支;自带轻量工作区备份,可一并还原文件(完整 Claude Code /rewind 语义)。 · DSH plugin: genuinely effortless in-window conversation rewind — never forking a new session; ships a lightweight workspace backup that restores files together with the rewind (full Claude Code /rewind semantics).
Delete sessions of a DeepSeek Harness profile from the UI: header danger button + session-row menu item with a risk-consent dialog; host endpoint + tool delete the log, projection cache and workspace accounting. Works in web and the desktop client alike.
DSH zero-dependency toolkit collection: time / encoding / json / calculator / csv / regex / markdown / diff / stat / schema in one entry point
Stable, editable PPTX generation for AI agents — semantic IR in, native DrawingML out
Tavily settings backend for the DeepSeek Harness: usage endpoint, flat key-list manager with usage strategies, and the web_search on/off toggle.
Give DeepSeek (text-only) models vision in Claude Code, Codex, and Agent Plugins clients: describe images via any OpenAI-compatible vision endpoint.
DSH peer link — point-to-point messaging between dsh agents and other local agent sessions (e.g. Claude Code) over unix sockets. Independent plugin: register as a peer, receive messages into agent context, reply with peer_send, list peers with peer_list.
Stable, editable PPTX generation for AI agents — semantic IR in, native DrawingML out
Agent2Agent (A2A) protocol v1.0 plugin for DeepSeek Harness (DSH): outbound client (remote AgentCard skills as ctx.tools), inbound server (A2A JSON-RPC + AgentCard endpoint), and a settings dashboard of live A2A connections
DeepSeek Harness plugin: quiz-style reading (答题式阅读) of any Markdown document — 5-stage game loop (diagnose→level→guided reading→checkpoint→settle) with a Web sidebar panel and a reading-area tab.
File-diff visualization between checkpoint time nodes for DeepSeek Harness: a read-only timeline + per-file line diff over the dsh-checkpoint-rewind checkpoints storage domain, with preview-first rollback (restore workspace files from any time node, single-file or whole-node; per-file current→snapshot preview diff; single-shot in-process undo), exposed as /diff and /rollback commands, a JSON HTTP API, and a GUI panel (session header action + shell overlay). Requires the upstream dsh-checkpoint-rewind plugin (the snapshot producer) to be installed first — without it, only the Trace scope works
DSH web plugin: roll back to before any of your questions, or edit the last one and resubmit, restoring the workspace files the discarded turns produced.
DSH plugin: a simplified 9router (AI routing gateway). Adds a 路由系统 (Routing System) sidebar entry beside 记忆系统; the panel shows suppliers, account pool, combos, and endpoint/API keys.
Track Bridge: embedded task-management engine for DeepSeek Harness — decision-point protocol, capture wall, and Linear-shaped issue store over session events + storage KV
Application-level Vision-Language-Model (VLM) analyzer for DeepSeek Harness: analyze_image tool with primary/backup OpenAI-compatible endpoints, automatic failover, and an auto-saving web settings page.
DSH Web client plugin: vertical message pointer rail on the left edge of the chat area, with clickable markers per user message for quick navigation and hover preview bubbles.
Exa Search MCP for DeepSeek Harness: mounts the remote Exa MCP endpoint (https://mcp.exa.ai/mcp) through the in-box @deepseek-ai/dsh-mcp-client bridge
Context compression tool (context_compact): the agent writes the replacement checkpoint itself and hands it to the host compaction engine, which skips the LLM summarizer call; automatic compaction stays on the official engine.
Native DeepSeek Harness plugin: gates every pwsh tool call against the PowerShell pitfalls from the blog post 'PowerShell 实战踩坑大全' via the official tools/pre-execute interception point, and bundles the powershell-check skill.
Host-level vision bridge for text-only models: analyze_image tool (Ollama local / Xiaomi MiMo cloud / any OpenAI-compatible endpoint) returning structured evidence.
Give DeepSeek (text-only) models vision in Claude Code, Codex, and Agent Plugins clients: describe images via any OpenAI-compatible vision endpoint.
Expose a DeepSeek Harness agent over the Agent2Agent (A2A) protocol: agent card, JSON-RPC endpoint, LLM-powered replies, and a live status card.