124 results
Structured context for DeepSeek Harness: split the conversation into typed sub-contexts (reasoning / tool calls / replies) and index them with a table of contents. 结构化 dsh 上下文:按类型拆分小上下文(思考/工具调用/正式回复),以结构化目录为索引。
Codex-style in-place conversation rollback and edit-and-resend for the DeepSeek Harness web GUI, plus a right-edge outline rail (session table of contents) whose search pages backwards through history until it matches
Right-to-left text direction for the DeepSeek Harness Web client. Marks Arabic, Hebrew, Persian, Urdu and Syriac content as RTL by script dominance, leaving Latin content untouched.
Personal usage analytics & activity dashboard for the DeepSeek Harness (dsh) web GUI: token totals, session activity, GitHub-style contribution heatmap, model / reasoning-effort / tool / skill / plugin rankings, streaks and insights. Aggregates real session events through ctx.sessionPersistence; no dsh core changes, no prompt content leaves the machine.
Keeps long coding-agent sessions lean: wrap work in named tasks and, when one is done, fold its whole span into a short titled summary. The conversation stays readable, context costs stay low, and every fold's original content can be read back on demand. For DeepSeek Harness.
Right-side drawer host for Web: every registered entry becomes a full-height drawer with its own edge tab (auto-stacked and group-centered), width resizable. Content plugins register via the dockPanels client registry. 右侧抽屉宿主：dockPanels 注册表、边缘标签堆叠、宽度拖拽。
Indirect prompt-injection guard for DeepSeek Harness: taints tool output by origin and gates privileged tool calls that follow untrusted content
Easy-fold plugin for the dsh Web GUI: sticky Think fold row and fold scroll compensation for long expanded content
Content-addressed proof that recorded DSH authorization decisions and effects agree
DeepSeek Harness plugin: content-routing LLM adapter. Text-only requests go to DeepSeek; requests containing images go to a user-configurable OpenAI-compatible vision provider. Read_image works natively for text-only models.
Installable chat content renderer extensions for DeepSeek Harness
Search field for the DSH Web Settings panel: independent candidate list over the content column, click to navigate to the owning section and focus the setting
A paper-review panel for the DeepSeek Harness: converts PDF papers into structured, selectable text (columns, headings, tables, sub/superscripts, table of contents) with figure and display-equation rendering, a floating ask-about-the-selection dialog, and per-project paper stores under a Paper workspace.
Content-free request and stream diagnostics for DeepSeek Harness
Offline content-addressed proof for supplied license-obligation delivery closure
Offline content-addressed proof for supplied access-review closure evidence
Offline content-addressed proof that independent builds reproduced byte-identical specified artifacts
Multi-source web search tool for DeepSeek Harness — queries DeepSeek / DuckDuckGo / Exa / Google / Bing in parallel, merges results, fetches page content.
DeepSeek Harness (dsh) plugin: export Markdown pipe-tables found in conversation content into Excel (.xlsx). Ships a host tool plus a Web client bundle.
Offline content-addressed DSH model-visible tool surface conformance evidence
Offline content-addressed proof for supplied vulnerability-remediation closure
DSH web plugin: full-text content search for the better-sidebar file manager — a search tab that greps file CONTENTS in the session working directory and opens matches in the sidebar editor
SQL 优化模式：EXPLAIN 分析、索引策略、N+1 解决、查询改写。受 wshobson/agents（38k★ MIT）启发。
dsh-skill-editor: edit skill SKILL.md content (body + name/description) right from the DSH web Settings page. Local skills edit freely; bundled / built-in skills are read-only with a warning.