Plugins
Durchsuche, filtere und installiere DeepSeek-Harness-Plugins.
750 Plugins gefunden
dsh-claude-project-ui
lanechenhku/dsh-claude-project-ui
Claude-style project UI for DeepSeek Harness on top of dsh-threads: one board gathering every project with its open threads and tasks, extra folders mounted per project via the native folder chooser, a bilingual zh/en panel with in-app settings, and thread reports rolled into tasks.
dsh-grill
kingzhoudk/dsh-grill
Per-conversation Grill switch in the composer tool row (also /grill on and /grill off): while it is on, the agent analyzes every request for ambiguity, missing information and contradictions, asks focused questions with 2-4 concrete options and a repair proposal, and plans or executes only after every blocking question is resolved. Trivial work is done directly with a one-line reading; non-trivial work is planned and stopped for approval. State is folded from the session log, so reload, resume and fork restore it, and a delegated subagent reports unresolved questions instead of asking.
dsh-vpn-guard
huanlin-labs/dsh-vpn-guard
Provider-agnostic egress monitor. Reports exit IP, country, ASN and latency, plus a per-service view of how that IP looks to AI services; an optional Clash or Mihomo adapter adds node lists, switching and traffic, and it never reads subscription links or credentials.
dsh-galatea (project-tasks)
loveandbeepbrother/dsh-galatea
Talk to one assistant and let your projects do the long work. Tell your Galatea assistant what you need; it hands the job to that DSH project's own session, which works in the project folder and reports back. The assistant checks the report against what you asked for (accept, send back for changes, or cancel) and tells you how it went. If the project is linked to a chat group, progress is posted there too. Requires the Galatea framework.
dsh-datapilot
xiduo-wang/dsh-datapilot
A persistent DSH sidebar workbench for local data analysis: nine chart types, column profiles, ONNX model graphs, uploads, structured reports, and built-in ML/DL training across multiple workspaces. Its HTTP routes have no DSH access check, so a web page open in your browser could call them.
dsh-siyuan-export
snaiheihei/dsh-siyuan-export
Exports a single dsh assistant reply into a SiYuan notebook as a structured Markdown note, with sections split by heading, icon-tagged bullet points and code blocks kept as-is. A button on every assistant message opens a dialog with a model-generated editable title and a document-tree picker, or runs silently and reports the result in a toast. SiYuan only and export-only, with no sync back.
dsh-super-leader
rickt34/dsh-super-leader
Leader agent preset that clarifies requirements, dispatches tasks to workspaces and sessions, and relays worker reports.
dsh-plugins (dsh-project-report)
skillpie/dsh-plugins
Project reporting in the DSH sidebar: draft a per-project report from its git commits and session records with one click, run it on a built-in schedule with a configurable daily cutover offset, and optionally submit Moka overtime automatically; all sessions of a project share one report.
dsh-novel-agent
good-saltfish/dsh-novel-agent
Right-sidebar novel studio that treats a chapter as a pipeline with checkable artefacts instead of one long prompt: scaffold a project, inspect the exact context pack a chapter gets, generate 2-6 parallel candidates and keep the one you pick, then run chapters in sequence with per-chapter progress on disk so it survives an engine restart. Length band, cross-chapter Jaccard and drift ratio are enforced in code rather than by model self-grading; rejections come back as repair evidence for up to two rewrites. AI-fingerprint metrics are reported against a 3956-chapter human baseline but never block, because the distributions overlap. Ships 7 agent tools. The writing engine is a separate Python package that the plugin installs on first use. Needs Python 3.10+ on PATH (or a path set in the plugin settings).
dsh-atan-daily
artaniszhou/dsh-atan-daily
Scheduled DSH tasks: each topic keeps its own folder and task-detail file, runs by cron or manually, and renders a timestamped PNG or PDF report that can be pushed to Lark.
dsh-fintax
fleekelon/dsh-fintax
Chinese tax and cross-border e-commerce workbench: a DSH sidebar workbench and settings view (sidebar.panellist, main, settings.section), a local BigInt statement parser for 8 platforms/wallets (Amazon, TikTok Shop, Temu, AliExpress, Shopee, Shopify, PayPal, and Stripe) CSV/TSV/UTF-16/XLSX/ZIP exports with multi-currency consolidation, rolling 12-month 75M/5M CNY screening indicators, and HTML/Markdown/CSV report export (fintax_parse_statement), a 10-scenario deterministic tax and reconciliation calculator (fintax_calc), 4 read-only Tencent IMA knowledge-base tools with retry, 60s read cache, and short-query fallback (ima_list_kbs, ima_browse, ima_search, ima_read), and 4 bundled compliance skills. Raw statements are parsed locally in memory via host-authenticated /api/fintax routes (/parse, /calc, /history, /status, /credentials, /kbs, /config); optional monthly KPI-only summaries may be archived locally in ~/.config/ima/fintax_history.json (0600, never raw order rows), and optional IMA credentials from env vars or ~/.config/ima/ are write-only (never echoed back) and sent only to ima.qq.com; pure ESM with zero runtime dependencies.
dsh-tiedan-update
hochan666/dsh-tiedan-update
Adds two rows to the DSH general settings section. Version update asks the npm registry for a newer @deepseek-ai/dsh, installs it as a background job and streams the install output into the row; restart DSH stops the running process so its supervisor can start a fresh one. A second restart request is refused, and a systemd unit that reports no main process is refused rather than respawned beside it. The row also reports the process uptime and identity, so a page reconnect can be told apart from a real restart.
hermes-skills-sync
hochan666/hermes-skills-sync
Imports skills from a local Hermes Agent installation into DSH. It scans the Hermes home directories and each profile below them, lists every skill with its description, file count and size, and copies a selected skill into the DSH global skills root or into one project, keeping the original folder name. Only a frontmatter name that DSH would reject is rewritten, an existing skill of the same name is never overwritten without a choice, and several skills can be imported in one batch with a per-item report.
dsh-provider-doctor
tangyixiao/dsh-provider-doctor
Read-only diagnosis for a model provider that stopped working. Walks the connection chain one link at a time — host runtime and profile identity, the plugin row, whether the provider route is registered, whether it is configurable, the credential metadata the provider itself reports, the model catalogue, the provider's own diagnostic and the last recorded request — and labels each link observed, inferred or unavailable rather than guessing. Diagnoses the `workbuddy`, `workbuddy-ai` and `zcode-cli` routes from their plugins' public diagnostic commands, and adds a sidebar panel plus a copyable redacted report. The offline pass sends no model request, starts no process and reads no credential file; the connection check is a separate opt-in the panel describes before running it. Version 0.1.10 applies the synchronous Host request policy followed by owner admission to all five HTTP routes before diagnostic actions; missing, throwing, malformed or asynchronous admission services fail closed. Declared hosts are DSH 0.2.2-rc.1 and 0.2.2-rc.3. Version 0.1.11 has been verified with DSH 0.2.2-rc.5.
ai-terminal-pet (dsh-ai-terminal-pet)
jjjjjjjjnnjnn/ai-terminal-pet
A floating always-on-top overlay for Windows that shows which of your terminals is waiting: it aggregates DeepSeek Harness, Claude Code, OpenCode and Antigravity session states (working, waiting for approval, failed, finished) over a named pipe, rings only on real events, and a click jumps to that terminal or copies its project path. Harness sessions appear on install; the other three need their own reporter from the repository, and approvals still happen in DSH.
dsh-worklog
ayf-master/dsh-worklog
Turns DSH session logs into Markdown work reports for a day, an ISO week or a month, grouped by working directory. Each session lists its requests, changed files, tool counts and interruptions, and a closing section names the most-changed files, generated documents and key commands. Adds the /worklog command, a worklog_generate tool and a settings page for the output folder, cross-day merging, the products list and an optional weekly timer.
dsh-quality-up
quality-up/dsh-quality-up
Tracks manufacturing quality issues and 8D improvement projects, analyzes process capability and SPC control charts, and exports reports from a local SQLite database.
dsh-clean-chat
ziye1208/dsh-clean-chat
A deliberately context-free chat panel in the right sidebar: it sends only a one-line system prompt plus what you type there, with no tools, no session history and no fork of the current conversation, and keeps the panel history in memory. Each reply reports its own token usage, plus the cache hit rate on routes that return cache stats.
dsh-migrator
fenshal/dsh-migrator
Lists every plugin installed in the current DSH profile, lets you tick the ones to keep, and exports them either as a migration sheet for a new machine or as a self-contained zip (local plugin sources plus a restore script) that a fresh DSH can install in one run. The restore script is PowerShell; a DSH restart (plus pnpm install if the plugin page reports pending installs) is needed afterwards.
dsh-better-debug
leonlee114514/dsh-better-debug
Appends one short user message before the next step when the model is circling: the same absolute path written three times with no measurement landing in between, the same error episode (tool, target and normalised fingerprint) returning twice, or a write followed by two steps with no measurement — the last one only arms, and the message lands before the next write rather than interrupting reading. A measurement has two states: intent is judged from the tool arguments at call time, and the write streak is cleared only when the result comes back, only for that command directory. A failed test still counts, because it yields the line that is wrong. Hard measurements are `lsp_diagnostics`, `verify`, `run_tests` and `run_test`; `ops_bash`, `bash`, `shell` and `pwsh` count only when the command matches a measurement pattern, so `ls` no longer launders a blind edit streak. Signals fire once per episode, re-arm after a real measurement, and are capped per turn; a successful `present` sleeps the gate until the next write, and subagent sessions are skipped. Generated paths (build output, lock files, logs) are excluded. Error targets are read out of real `file:///` ESM stacks, so two different files never collapse into one episode, and `cd <dir> && cmd` counts as a measurement of that directory rather than its parent. Ships the `stuck-gate` skill at bundled rank 600 (a user-level copy at 400 overrides its wording) plus a read-only observation panel at `/better-debug` reporting firings, sessions covered, the first action after each message across six classes, and how many measurement loops actually closed. The log defaults to `$DSH_HOME/logs/better-debug.log` and is created on first write, so the panel has data right after install. Optional host services are subscribed dynamically, so a host without them still loads the gate. It does not estimate tokens saved, and it does not claim the message caused the change. The panel banner counts only the failures after the most recent load (`log.current`), so one old mount failure no longer reads as "currently disabled". Test suites wait for log conditions instead of fixed delays (test-only change in 0.2.5).
dsh-skill-manager
hightoward/dsh-skill-manager
Skill manager panel for DeepSeek Harness: scans the six workspace and user-level skill roots, files every skill into fifteen built-in type cards (and any types you create yourself), and lets you move a skill between types without editing its files. A model you pick writes a one-sentence Simplified Chinese description for each skill, and can draw the icon for a custom type from the type name and the skills placed in it. Also reports which skills name each other in their bodies, and checks installed versions against SkillHub and upstream git sources.
dsh-scihub
stallenlu/dsh-scihub
Adds four tools: scihub_pdf / scihub_batch fetch papers by DOI or article URL from Sci-Hub, reporting per-DOI status honestly; libgen_search / libgen_download search Library Genesis by keyword and fetch a file by MD5, keeping its real extension (pdf / epub / djvu). Files are named from citation metadata via a configurable fileNameTemplate and written atomically without overwriting. All requests support an HTTP/HTTPS CONNECT proxy (config.proxy or env vars, with NO_PROXY). Batches run with bounded parallelism, skip dead mirrors, and reject truncated or verification-wall responses.
dsh-turn-cleanup
leonlee114514/dsh-turn-cleanup
Blockiert das Turn-Ende über agent/turn-stopping bei verbliebenen temporären Dateien: löschen oder einzeln begründen und sinnvoll verschieben. Prüft .agent-tmp/ im cwd sowie in diesem Turn über fs-Hooks erfasste, per realpath normalisierte Pfade im cwd mit präzisen Temp-Mustern. Git, Abhängigkeiten und Builds sind ausgeschlossen; alter Müll wird nicht erneut beanstandet. Überzählige .bak berührter Dateien werden standardmäßig gemeldet. Ein Gate je Turn, keine Unteragenten. Löscht nur bei explizitem Backup-Modus prune; Text folgt der Host-Sprache.
dsh-ledger-wealth
ainachoneko/dsh-ledger-wealth
Lokale Buchführung mit separatem Vermögensregister: Texteingaben, Beleg-OCR über einen optionalen lokalen Anbieter, laufender Saldo, strikt sinkende tägliche Gewinnfolge und zweigeteilter Monatsbericht, der Bargeld und Vermögen nie zusammenzählt.