Plugins

Nai milk-frog mascot appearance layer for the DeepSeek Harness web GUI: wears the bundled milk-frog artwork as the GUI background (surface-token fade with opacity, blur, scrim and position controls), seats the matching square avatar as the sidebar and conversation brand mark, and edits it all from a Settings section; the host serves the artwork and config locally under /api/nai, and the config lives at $DSH_HOME/nai.json.

Windows anime companion for DeepSeek balances and Codex quotas, with integrated Codex multi-account OAuth, per-account quota and model management, compact mode and autosaving settings. Optional panel account cycling and separately authorized daily warm-up; multi-account startup/5h-full triggers remain unavailable.

Detect DSH plugin conflicts and loading failures, quarantine affected bundles and recover them, with offline startup rescue.

Adds a Hindsight configuration and status dashboard to the DSH sidebar for connection, model, embedding and bank settings, with health checks and ownership-checked local service controls; memory tools are provided by the separate official integration. Hindsight itself is a separate service and is not installed by this plugin.

Exposes the Paper-gogo evidence-first paper workflow as native tools: all 18 phases and the seven G0-G6 gates become callable, gate results are normalised and advancement is refused while any gate fails, and claims are audited against the four evidence states SUPPORTED, INFERRED, VERIFY and MISSING. Ships the workflow text and 57 skills in its own directory, so it installs without a separate Paper-gogo package.

Pops the DSH side chat into a frameless, always-on-top Windows desktop window that can sit over any other application, with its own settings page for appearance, pinning, remembered geometry, opacity, ephemeral threads and cache cleanup.

Read-only Skill inspector for DeepSeek Harness. It shows which Skills a conversation actually loaded — from observed loading evidence only — and which Skills the current environment can discover, and gives each Skill a detail page whose left column switches between the framework, the evidence, validation, evaluation, step evidence, run logic, the before/after of a modification when there is one, the definition facts and the raw SKILL.md. Declaration and run stay apart: the body is split into structure, declared workflow and progressive disclosure, the run logic is read from local receipts, and every declared step carries its own evidence — no observed evidence is reported as exactly that, not as executed or not executed, and the run's instruction hash sits beside the current definition as match, mismatch or unavailable. The evidence view gives each subject one of three states (declared, observed, unavailable), walks definition → load → use → outcome, names what each stage cannot reach, declines to attribute an outcome to the Skill, and can copy the snapshot out as byte-stable JSON. Validation checks the Agent Skills open standard and the platform profiles separately, each finding on a rule id and labelled standard compliance or platform compatibility. Evaluation keeps repeatable cases with the conditions each run was recorded under and the four stages of runtime evidence, comparing two runs of one case item by item without scoring them. The trigger boundary keeps should-trigger and should-not-trigger prompts, has you run them in a real new session, then records what was observed — whether a load record appeared and whether it was model-invoked or user-explicit; a typed /name is not a natural trigger, and an absent load record is reported as not observed, not as not triggered. The same projections are exposed to the agent as a read-only skill_review tool, which returns a structured review packet and says what it omitted when the packet will not fit whole. The inspector, the evidence chain and these reviews are computed on this machine from the definition and the session receipts; the Chinese reading version sends the body through the host's model when you ask for it, keeping code fences, paths and frontmatter verbatim and storing the result on this machine. A Skill can also be cloned into a local copy. It does not write back to the source Skill and does not run its scripts.

Cut LLM costs: a Jev (System One) classifier vets each turn against a declarative policy and recommends the subagent model route, so routine work runs at flash rates while the main model keeps the hard parts; explicit /route trigger, runtime delegation-capability gate, credential redaction, and offline policy replay over the verdict log.

DeepSeek Harness plugin that fixes RTL/LTR bidi rendering in chat: per-block direction for Persian and Arabic messages with embedded English, plus list markers and quote bars that follow their text

AMI MegaRAC OneTree 一键日志（OpenBMC dreport dump）上传与分析框架：设置面板「OneTree 日志」上传压缩包或登记服务器本地路径，安全解包、自动探测 dump 根、文件清单分类，为第二阶段逐域解析提供内核。

Enhanced desktop shell for DSH: auto-starts the local backend on launch (no browser tab), asks before closing the backend on quit, and starts the backend in parallel with Electron for faster cold start. Fork of dsh-clean-desktop-shell.

A dsh-better-sidebar reading view for .docx files: headings from the style map, paragraphs, lists, real table cells, inline pictures and header notes. Ceilings on archive size, per-part inflated bytes and rendered blocks keep a document from freezing the tab.

macOS desktop launcher with an installation skill: starts or reuses the local DeepSeek Harness web service and opens it in Google Chrome, with a transparent character and thought bubble icon.

Folds long composer drafts into a compact summary card once a character threshold is reached, without changing message submission.

DSH plugin that registers the ai-eight-honors-eight-shames skill: eight checkable working rules for coding agents (verify before calling, confirm ambiguity, cite evidence, reuse before creating, test before claiming done, follow local conventions, separate fact from inference, keep changes reversible) plus a ninth rule that logs each user request and the change made in two fixed report files at the workspace root.

Task-to-skill pairing meta-skill with a laziness ladder for minimal loading, a user-editable routing table, and a quarantine, SkillSpector scan, human-approval install flow.

Media-scape theme for the DSH web UI: standing/character-art and video wallpapers (mp4/webm, local + online) with boot animation, base color schemes (single-source JSON + presets + picker), capsule host-element styling, ambient particles, background music player (covers/playlists/random), typing sound effects, and a draggable dock toolbar with a unified upload entry — wallpapers and music uploaded together (suffix-classified), mid-upload no-disk buffering, pause→.part snapshots with offset resume, filename+size dedup with "skipped" notices, and true-delete removal with list refreshes.

0.8.6 fixes a startup crash: the host runtime package is an optional peer, never an installed dependency. TypeSafe Jev (System One decision model) wired end to end into DeepSeek Harness (0.8.6). Seven tools: jev_decision (1-25 parallel atomic verdicts in one call; noul/choice/score; typed answers with calibrated confidence; measured median 266-484 ms; verdicts only, never prose), jev_choose (rank 2-10 candidate approaches - fit 0-3 and risk - into an explicit composite with a recommended pick), jev_batch (NEW in 0.8.6: one set of 1-5 judgments over 1-20 texts, one real call per text at bounded concurrency, returning a paste-ready markdown table sortable by a judged dimension; built-in sentiment/priority/intent/pii/spam presets; a failed row is reported as failed with the raw error, never guessed), jev_usage (locally measured usage: calls, tokens, estimated cost, latency median and p95, per-tool mix, rolling windows, projected burn rate, optional local hard budgets, and a declared-balance remainder), jev_overview (session ledger), jev_guard_status (guard counters plus a registrations block that surfaces per-surface registration failures) and jev_verify (frozen 27-question live benchmark at concurrency 6: 96.3%, 26/27). WHAT IS NEW IN 0.8.6 - the ledger survives a restart, the remainder is derived instead of invented, and the plugin can now do batch paperwork. (1) The local ledger persists to $DSH_HOME/jev-usage.json BY DEFAULT, so the cumulative total no longer resets when the host restarts. Two real bugs behind the old silent zero were found and fixed: the ledger path only honoured DSH_HOME (absent inside the host process, so persistence was silently disabled and the file name was null), and save() wrote the still-empty in-memory ring over the history file because it ran before the first load while the debounce window swallowed the follow-up write. (2) Remaining is derived, never invented: TypeSafe exposes no balance endpoint - /v1/usage, /v1/quota, /v1/account, /v1/account/balance, /v1/balance, /v1/credits, /v1/billing, /v1/me, /v1/limits and a dozen more all return 404, and only /v1/models answers, with no balance headers - so the panel shows the remainder as your declared balance (quota.declaredBalanceUsd) minus locally measured spend since quota.balanceSince, and says plainly that it cannot be fetched instead of printing a number. (3) The composer pill re-reads every 20 s instead of every 60 s, so a call you just made shows up while you are still looking at it. (4) The collapsed pill leads with the remainder once you declare a balance, and the cumulative row is renamed when the ledger is on disk. Every tool has an in-conversation structured view; the Settings > Plugins card covers credentials, tool switches, the auto-guard, the dashboard, the prompt section, usage/quota and auto-triage. Honest by design: no API key means an explicit error, never fabricated results; every registration is quarantined so a failure degrades one surface. Tests 38/38 (npm test, ~2.5 s). Client half applies on refresh; server half needs a host restart.

Local SQLite long-term memory with CJK-aware search, BM25 and character n-gram cosine reranking, deduplication, and lifecycle management.

Animated DSH Web pet with model-adaptive output pacing, tool-state reactions, replaceable character packs, and a bundled skill for creating or extending pet animations.

Speaks your agent's reports in a voice you choose, compressing long reports before speaking. Character voices need no training - a 3-10 second reference clip clones one, and community-trained models work too - and no 6.4 GB GPT-SoVITS install: the plugin installs its own runtime and voice. An existing GPT-SoVITS can be used instead, and it is the same voice model.

Automatically highlights key points, definitions, warnings, and questions in completed assistant answers using one separate model call.

Pre-install compatibility gate for DSH plugins: locks the resolved version, checks peer/engines declarations, dry-installs into an isolated DSH_HOME and boots it with a timeout, blocks on any failure, and restores the profile if the real install fails. Optionally intercepts dsh-market install/update clicks. Compatible with DeepSeek Harness 0.2.x.

Plays a voice line or a sound when a turn finishes and a failure cue when a tool call or a turn errors — user aborts stay silent by default and subagent sessions are skipped, both configurable. Three alert modes: voice, effects, or off. Effects is the default and ships 20 built-in sounds (10 alert chimes + 10 nature), each assignable to the done, fail or approval event with per-row preview, so it needs no API key and no audio files; the voice route uses Volcengine TTS with cloned or preset voices, batch-imported voice IDs and fingerprint reuse that skips re-synthesis when text, voice and parameters are unchanged. A third approval slot exists but has no automatic trigger — it only plays when called manually. Playback goes through winmm/waveOut as-is, never changing the system volume or mute state, and the player is spawned detached so session teardown cannot silence it. Windows only.