520 results
File-change review for DeepSeek Harness sessions: Codex-style per-turn cards (one per turn with workspace-relative paths), a turn-scoped review tab with syntax-highlighted diffs and sticky file headers, per-turn revert via op sequences, light/dark color presets that follow the DSH theme live, live polling — powered by the official session-history channel.
Host-level vision bridge for text-only models: analyze_image tool (Ollama local / Xiaomi MiMo cloud / any OpenAI-compatible endpoint) returning structured evidence.
DSH 中转站模型思考强度配置与自动识别插件
DeepSeek Harness desktop hub — Tauri 2.x shell + dsh web plugin layer. This npm package ships the plugin half (client UI + HTTP routes + cordis.patch.yml); the desktop shell ships as the NSIS installer (see build/).
Give DeepSeek (text-only) models vision in Claude Code, Codex, and Agent Plugins clients: describe images via any OpenAI-compatible vision endpoint.
DSH tool-call compatibility: strips sandbox_permissions/justification from tool-call arguments for third-party models (GPT etc.), and lets the user skip possibly-stuck tool calls with an LLM-facing notice.
A Symphony-compatible multi-provider task orchestrator and dashboard for DeepSeek Harness
DSH plugin: renders images inline in DeepSeek Harness Web chat and gives text-only models vision — cloud multimodal API first, local Qwen3-VL fallback.
Sidebar file tree for DeepSeek Harness: a files panel with a lazy, per-extension-icon tree, copy-path and send-to-agent actions
DSH PDF 解析插件: 为 agent 提供 pdf_info / pdf_extract_text / pdf_render_page 三个工具。基于 pdfjs-dist + @napi-rs/canvas; 对未内嵌字体的 PDF (如嘉立创EDA导出) 自动改用系统字体渲染文字。热插拔 — 通过 cordis.patch.yml + profile node_modules 软链接挂载, 无需改动 dsh 源码。
Cross-session memory with governance, self-evolution and an approval-gated write path for DeepSeek Harness. Heat-ranked recall, environment conventions, forgetting lifecycle, supersession chains, poisoning defense, auto-distillation, and an open measurement ABI.
A semantic, theme-compatible motion runtime for DeepSeek Harness.
Enhanced subagent delegation tools for DeepSeek Harness: per-call model/provider/persona/toolFilter overrides, @preset: references, and provider/model composite ids — zero patches, bundle.
DeepSeek Harness plugin: hide the raw chain of thought and show a summarized chain of thought produced by a small model (any Chat Completions-compatible API).
Eyes for text-only DeepSeek on DeepSeek Harness: image understanding via Antigravity IDE quota (default, flash/pro), OpenAI-compatible VLM endpoints, Gemini API, or local Ollama — model-invokable vision tool, wrapper adapters for deepseek/opencode-go, evidence memory, and a polished client panel.
AgentAstra — a spatial agent-trajectory view for DeepSeek Harness
DSH compatibility plugin that treats redundant non-escalating sandbox permission fields as a no-op
dsh bundle: attach any file type in dsh web — text-likes inline, binaries as workspace path references, rasters via the built-in pipeline
DeepSeek Harness adapter that exposes each configured vision model as a selectable DeepSeek composite in Web UI.
Enhanced subagent delegation tools for DeepSeek Harness with per-call cwd control: everything in dsh-subagent-tools plus a cwd parameter, shipped with the two in-process provider patches it requires.
Cache-stable progressive tool discovery and dispatch for DeepSeek Harness
dsh-polyglot — the model switch for DSH: one generic OpenAI-compatible ctx.llm adapter, curated free/cheap DeepSeek provider presets, and automatic provider fallback when a free tier rate-limits you
DSH JSON Schema validation tool: validate, failure paths, static explain, default-only normalize; zero-network, zero-dynamic-code with resource limits and a pattern ReDoS worker budget
DSH regex tool: test/find/replace over text plus static pattern explanation, zero-dependency with ReDoS guards