Plugins

Share a selected chat segment as Markdown, HTML, or plain text; Header button, sidebar ... menu, and /share command.

Web UI plugin that folds each staged response's think and tool calls into per-segment bounded, scrollable slide windows with a configurable height, keeping every intermediate response visible.

Rotates the model status line through a 53-phrase English/Chinese deep-theme pool: a random pick at turn start, event-driven switching on new thinking segments or tool calls (3s throttle), and a click on the status line toggles the language.

Translation toolkit for DSH agents: sentence segmentation, term and glossary extraction, source-target QA (numbers, units, brackets), tone guides, and a durable workspace translation memory.

Adds a Share to chat action under each reply: converts the answer to WeChat or WeCom text, ordered segments or a long image, editable-table email HTML, or a downloadable Word document.

Pre-implementation design review gate: an Elephant agent drafts the design, zero-context Goldfish subagents attack it from fixed reviewer roles, and you approve before implementation.

Local long-term memory for DSH sessions: archives older turns under context pressure, with four tiered hot-memory boards, project documents, memory families and root transfer, manual injection, 20K segmented reads, and tag-related retrieval. All data stays in the workspace.

Lets build commands run inside the DSH sandbox instead of asking for elevation on every invocation: redirects pure caches (GOCACHE, NPM_CONFIG_CACHE, BUN_INSTALL_CACHE_DIR, PIP_CACHE_DIR) to a temp root, or grants the real cache directories write access in place so they stay shared with the terminal. Dependency stores (GOMODCACHE, CARGO_HOME) are never redirected, only granted. Adds optional user-directory grants (~/.config and friends, all off by default) and a command allowlist that matches exact strings or per-segment after splitting on &&, ||, ; and |, while rejecting redirects that write files (fd duplication like 2>&1 and /dev/null are allowed), command substitution and backgrounding outright. Settings are per-profile or per-session; file edits apply without a restart.

Bundles 50 AMD ROSS agent skills for Vivado, Vitis HLS, Versal segmented configuration, Vitis AI, and hardware debug into DeepSeek Harness, with a left-sidebar catalog panel for filtering, searching, and copying skill invocations.

Context rollover: auto-saves a handoff before the context window fills and re-injects it once in the next session segment. Two tools and a settings panel.

A personal usage profile page in the DSH sidebar. Five headline numbers (cumulative tokens, today's tokens, longest task, longest and current active-day streaks), a 12-month activity heatmap with daily / weekly / cumulative scopes, a Cache Lens that splits every token into cache hits, uncached input and output as a segmented bar (with a zoom into the uncached split), a stacked bar chart of tokens over time grouped by model family (version plus tier, so vendor-specific ids and unversioned aliases collapse correctly), and a per-provider split for today or all time. A page-level range (all / today / 7 days / 30 days) drives the token figures. Everything is folded from this machine's own session logs through ctx.sessionQuery. The only outbound request is a six-hourly read of this package's own version from the npm registry, used by the in-panel update button — no telemetry, no credentials, nothing uploaded.

Full-screen boot animation for DeepSeek Harness: a wave sweeps the screen, a whale assembles from particles inside three breathing dot rings, and a four-language salute lands over a segmented progress bar; ships an 18-second cut and 3.8-4.5-second short cuts, a language picker in settings, and a four-language preview queue.

Dictation that ends up in the repository instead of the chat log: hold the button (or import a WAV), and the plugin archives voice-memos/<id>/ with audio.wav, a rendered waveform.png, memo.md (frontmatter, per-segment timestamps, ready-to-paste quote blocks) and meta.json. Audio is resampled to 16 kHz mono in the browser before it is sent. Transcription stays local — the harness speechToText service with silence-split windows, or a command worker such as faster-whisper — and when no engine is available the memo is still filed with an explicit status and reason rather than a fabricated transcript. Ships a conversation-view 口述 tab with click-to-seek playback and per-segment quote copy, three agent tools (voice_memo_save / voice_memo_list / voice_memo_read) and HTTP routes. Zero runtime dependencies, 52 node:test cases, real screenshots in docs/.

Outbound route pool for DeepSeek Harness. Ships 15 seed routes plus a 20-entry candidate pool covering 7 target kinds (GitHub raw/codeload/main, npm registry, PyPI index, HuggingFace, general). Routes are verified before use: upstream IP lists go through a TLS handshake with full certificate-chain validation (a bare TCP check can be answered locally and report dead IPs as healthy), and failures advance a cooldown ladder of 2/5/15 minutes with the fastest healthy route picked per target kind. Routes are applied by rewriting the URL with one of four rules (prefix, registry, index, host-replace), reaching git, pip, npm, pnpm, gh, curl and wget through three paths: the shell seam (download commands the model runs through the bash tool are intercepted and rewritten automatically; commands wrapped by the pwsh tool, and fetches made by the host itself or the renderer, are not covered), two registered tools the model can call directly (line_pool_status, line_pool_run), and child-process spawn. Multi-line commands keep their line breaks and only the download segment is touched. URLs carrying credentials (user:token@host) are never rewritten, so tokens are not forwarded to third-party mirrors. No global config is modified. A sidebar panel lists every route with latency, failure reason and cooldown countdown, supports per-route retest and enable/disable, and shows live TUN/proxy status from network-adapter and loopback port checks. The candidate pool is re-verified every 24 hours so newly usable routes are added and dead ones dropped. 570 tests, 87% line coverage.

System One decision runtime for dsh. Ships fast non-generative tools (choice, score, screen, gate) on a local or hosted System One gateway, and — without the model calling anything — prunes oversized tool output segment by segment, refuses repeated identical calls once System One confirms stagnation, verifies completion claims against real tool evidence, and corrects failing retries. Published on npm as dsh-harness-reflex (plugin) and harness-reflex (core, also a stdio MCP server).

Startup video intro with per-clip playback parameters (segment, speed, brightness, fit, blurred cinema fill, hold ratio and fade duration) and light/dark clip variants selected by the app theme.

Codex-style reasoning-effort card for the composer: takes over the model selector and shows the current level, the model row and a segmented slider over the model's own adapter-provided levels, with a model list, a reset, and a purple aurora with twinkling particles at the maximum level.

Ships a skill for verifying a change with the cheapest check that still yields a definite answer: an L0-L3 cost ladder from static checks to a single suite, condition polling with a timeout instead of fixed sleeps, assertions that can actually fail, and durations that are measured rather than remembered. The package also carries the two zero-dependency Node tools the skill calls - wait-until.mjs (poll a command, a file, or a port until it is ready) and measure-runtime.mjs (wall-clock timing for any command, plus a segmented timer for library use).

Context compaction that keeps its receipts: a five-section handoff summary, a durable segment index, review tools that read compacted text back, optional working notes, and a read-only extraction exit. Mounts as the compaction-basic row's replacement; cha

DeepSeek balance and per-session token usage in the session header, with a three-level balance alert and editable thresholds, one-click top-up links, a segmented token bar with cache-hit statistics, and a visibility-aware poller.