Plugins
Browse, filter, and install DeepSeek-Harness plugins.
2731 plugins found
dsh-im-indicator
poy-coder/dsh-im-indicator
Puts a channel capsule and a live status dot on the one session each IM channel is currently driving in the sidebar; zero dependencies, about 7 kB. Its HTTP route authenticates every request through DSH connection.requestRejection, so a foreign origin receives 401/403 with an empty body.
dsh-plugin-litsearch
peijiping/dsh-plugin-litsearch
Model-driven, multi-round literature search in chat: the sidebar visualizes results, rounds and full-text availability, across ten free sources — PubMed, Europe PMC, OpenAlex, Crossref, Semantic Scholar, arXiv, PubTator, DOAJ, OpenReview and OpenAIRE — with PDF download, citation export and Zotero from either side.
dsh-subagent-claude-code-pro
ckanner/dsh-subagent-claude-code-pro
DeepSeek Harness plugin: official Claude Agent SDK with your local runtime — per-call model and effort, durable sessions, and multimodal input/output.
dsh-clawbot
yunhai89/dsh-clawbot
WeChat ClawBot (Tencent iLink) channel for DeepSeek Harness: QR login in the DSH Web UI, one persistent DSH Session per WeChat user, conversation/model/workspace commands, and tool approvals forwarded to WeChat. It embeds Tencent's official openclaw-weixin plugin through an OpenClaw host compatibility layer instead of reimplementing the WeChat protocol.
dsh-workspace (dsh-plugin-airlock)
gabrielbav/dsh-workspace
Puts every plugin, skill and subagent bundle through 6 security scanners (Cisco Skill Scanner, SkillScan, Cisco MCP Scanner, OSV-Scanner, Trivy, Gitleaks) and an optional LLM judge on your own DSH model, in a quarantine with install scripts off, checking for prompt injection, MCP tool poisoning, vulnerable dependencies and leaked secrets; then it installs, asks you, or blocks, and confirms the installed files are the ones it scanned. Works on stock DSH.
dsh-wechat-draft-gate
dingchenhui0618-arch/dsh-wechat-draft-gate
Two-stage approval gate for outbound WeChat replies in DeepSeek Harness: every reply is staged as an editable local draft that keeps the quoted original and its voice-transcription source/confidence; only a panel confirmation or /send <id> releases it through a configured bridge, and unconfirmed drafts expire with a recorded reason. Dry-run by default, JSONL audit, zero dependencies. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-incidental-learn
oxmumu3/dsh-incidental-learn
Captures engineering and product trade-offs from daily Agent turns into a local skill ledger with collapsible postscript cards, automatic background staging, right-sidebar drills, and four-tier mastery checkpoints.
dsh-ai-study-kit
jerryjiao/dsh-ai-study-kit
Registers five user-invocable study-coach skills for DeepSeek Harness: ask-coach (learning-state probe plus the F1-F13 flows: bootstrap, new theme, daily study, wrong-answer recap, podcast, lesson authoring, content edit, verify, deploy, coaching, exam sprint, knowledge-graph projection, upgrade), study-coach, study-doctor, study-recap and study-podcast. Ships a copyable quiz-site template (kit/ snapshot, v0.28.0) that the bootstrap flow copies to ~/study-kit; every skill is slash-command only and does not participate in model auto-invocation.
residency-history-taking-internal-medicine
flying-ovaltine/residency-history-taking-internal-medicine
Residency OSCE history-taking trainer for internal medicine. It plays the standardized patient while you interview, then switches to examiner and scores the interview against the official 100-point rubric with itemized deductions; ships 32 internal-medicine case scripts and an append-only training log.
dsh-frontend-replay
zhan1024/dsh-frontend-replay
Frontend Replay records and replays frontend bugs, verifies fixes against explicit acceptance conditions, and exports browser evidence and Playwright tests.
dsh-opticforge
wbmgggg/dsh-opticforge
KB-first optical-engineering task assistant. opticforge_assist aligns a concrete problem with the 40-section / 324-entry knowledge base before any web search. 3.1.1 activates up to three topic playbooks (16 cards, 98 dimension declarations), retrieves evidence per required dimension, and returns a dimension-by-source coverage matrix, a gated plan with dimension IDs and cross-check steps, formula cards, a five-section report outline and acceptance checks. Answers explain principle -> formula/criterion -> approach -> steps, and never expose KB layer or entry IDs; sources are named as documents, standards, manuals or URLs. web_gate allows external search only for declared KB gaps, standard versions, vendor parameters or price timeliness. opticforge_verify checks KB alignment, present-only citations, dimension evidence, cross-relations, source tiering, KB conflicts, 30-day freshness and five-section presentation, and reports an adoption gate. Includes 308 concept cards, 6 generic dimensions, 19 formula cards and the opticforge-task DSH Skill. 308 entries are citable; 16 declared gaps remain, limited to A03, C01 and E01 verbatim sources. Five tools, zero runtime dependencies, no calculator, no network access and no action execution by the plugin. npm publication is on hold; install from GitHub with node scripts/install-local.mjs.
dsh-prompt-library
super-qaq/dsh-prompt-library
Prompt library for the agent and the composer: save prompts, search and filter by tags, insert into the draft or copy, import prompts from a link with the default model, and track skill or plugin dependencies.
Davinci_MCP_DSH_Demo
jiangxingchen4210/davinci_mcp_dsh_demo
DaVinci Resolve color console (DEMO): LOG restore with official LUT matching, 3-layer 15-node color graph, 10 sourced styles, delivery approval; macOS only.
dsh-plugin-trim
haidai233/dsh-plugin-trim
Trims the tool surface per session. Each session owns a profile — a tool whitelist — applied through `agent.ctx.tools.restrict({ allow })` on that agent scope only, and switching one session never touches another; `full` is the built-in no-restriction profile. The provider only picks the default: a session created or switched onto a local provider (explicit `localProvider`, `routes`, a loopback/private baseURL, or a local-ish id) gets the `local` profile, a cloud one gets `full`, and the re-plan runs on model-selection events plus one delayed pass because the projection can land after the event. Profiles are editable from the panel and persisted under `~/.dsh/plugin-trim-state.json`; an explicit per-session choice overrides the provider default until cleared. A separate, opt-in layer physically disables whole plugin entries through `pluginManager.setPluginEnabled` — that one is profile-scoped by API and the panel labels it as such rather than pretending it can be per-session. A `tool_trim` tool exposes list/status/profile/set/apply/restore to the model. The panel is served by a loopback-only HTTP server the plugin starts on its own fixed port — `http://127.0.0.1:12256/` (auto-bumps on collision, configurable) — because the DSH port changes on every restart and a locally linked package gets no market card in the sidebar; it lists live sessions with their profile and tool count, an editor per profile, a provider picker limited to providers declared in the profile or currently in use, and the full plugin list with search and paging. Two implementation details worth knowing: `tools.restrict()` is an intersection and cannot widen, so the old restriction is released before the visible set is read — otherwise switching to a bigger profile would shrink instead of grow; and every disposer is registered on `ctx.effect`, without which a hot reload leaves an unremovable restriction behind. Measured on a 12GB laptop GPU with a 27B model and a fully resident 256K cache, same conversation: 105 tools / 56,306 prompt tokens / 22.5 tok/s decode, 62 / 34,207 / 43.0 after disabling three tool-heavy plugins, 14 / 12,890 / 66.8 after a session moved to the local profile, with prefill rising from 487 to 1,310 tok/s. Restricting also drops the skill catalog injection, roughly 4.3K more tokens. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-workspace (dsh-plugin-aws)
gabrielbav/dsh-workspace
Gives every DSH session AWS's official MCP servers: 59 servers in 14 groups (documentation, IaC, EKS and ECS, Lambda, DynamoDB and Aurora, CloudWatch, cost and billing, Bedrock and more), switched on from a Settings page by group or one at a time. Versions are pinned, servers start read-only where they support it, and secrets stay in the DSH credential store.
dsh-split-proof
dingchenhui0618-arch/dsh-split-proof
Read-only split-rule acceptance for DeepSeek Harness: point it at a Clash/Mihomo profile plus a JSON suite of expected routes (direct / proxy / block or a concrete policy) and it walks the rules in their real order, reporting per domain the expected route, the matched rule with its index and raw text, and what actually happens. Rules that need DNS, an IP-geo database or phone runtime state are marked as undecidable offline instead of guessed. Optional baseline diff checks that domestic direct routes did not move. Ships an agent tool, a conversation-view panel, a CLI with exit codes, and a self-contained HTML report. Zero runtime dependencies, no network, writes no device configuration. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-github-access-guide
2005-626/dsh-github-access-guide
Sidebar panel that repairs GitHub access on restricted networks — a local racing proxy that reversibly takes over the Windows system proxy, a git accelerator rewrite verified against the native address, live channel status and a URL converter. It says plainly what it cannot fix: links that are intermittent, not addresses that are blocked outright. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-session-retitle
sirius2alpha/dsh-session-retitle
One-click session title regeneration for the current session: a labelled button beside the title, a row in the session header menu, the /retitle command, and a retitle_session agent tool, all calling the composed session-title service's explicit refresh.
dsh-plugin-models-fingerprint
darkhe55/dsh-plugin-models-fingerprint
Monitors DSH model routes by comparing protocol fingerprints and behavior statistics across sessions, with per-turn cards, route equivalence, official baselines, and drift/quality signals. Fingerprints are evidence, not proof of model identity. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-plugins (dsh-workbuddy-turn-credits)
yggizz/dsh-plugins
Shows what each completed turn actually cost, on that reply's action row: WorkBuddy sessions show credits, sessions on the official DeepSeek routes (account login or API key) show an amount in CNY. It measures instead of estimating — the matching balance is read at the start and again at the end of every turn, and the difference is that turn's charge — so each turn costs two read-only balance queries (no credits spent) plus about a second of settle time before the number appears. A switched account, overlapping requests, or an unsettled balance shows an em dash rather than a guessed figure, and a session that uses none of these routes shows no row at all. It reads the DeepSeek balance endpoint and the WorkBuddy credits route; the Host side answers only loopback requests. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-sidebar-density
pcysanji/dsh-sidebar-density
Sidebar density controls for DeepSeek Harness: row height, spacing, font size, logo size, new-session height, double-column plugins, plus patch-gated caption height/font and unlocked sidebar width. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-icon-changer
jiaxusummer/dsh-icon-changer
Changes the DeepSeek Harness desktop app icon (window, taskbar, tray): use your own image or restore the official set. DSH quits and restarts once to apply it. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-decision-engine
leonhardthust/dsh-decision-engine
dsh-decision-engine: a cheap judgement layer for DeepSeek Harness — a small decision model answers structured questions before the main model acts, so calls can be allowed, escalated, rerouted to a cheaper model, or skipped. All eight decision points (too
agentschat
swswordholy-tech/agentschat
Connect Claude Code to AgentsChat — AI Agent social network. Core tools stay lean while extended tool groups load on demand for lower token overhead and cleaner role-specific context.