Plugins

Approval frame for DSH tool calls: a rule table with four outcomes (deny, ask, AI review, allow) across tool, command, file and network scope, then AI fast/deep review and human approval, a hash-chained audit trail and a settings UI. Target files are inspected by metadata only (no file contents are opened); the reviewer sees the command, targets and clipped write payloads, with secrets redacted before anything leaves the machine.

Turns a stated rule into a checkable intake card and writes it into the right AGENTS.md, and keeps project plans, a rule ledger, project records, handoffs, a change log and a timeline.

Read-only workflow panel for DSH: overview, six gates, milestones, org template, task flow and roster - project data from a project.json file, roster from the live DSH state, ledger presence from the ledger directory. Ships as a Cordis bundle with two GET-only routes and writes nothing to the project or the DSH state file.

Skill shell that installs, runs and troubleshoots React-Monitor-Agent, a human-agent operation board built on review replay and checkpoints: the agent writes its work as a structured replay timeline of screen-recording, screenshot, diff and text nodes, and the human decides at plan, progress, risk and delivery checkpoints, with verdicts branching the timeline and rejected branches kept greyed-out for audit. The skill guides cloning the source repo and starting a zero-dependency Node event service on 127.0.0.1:8790 plus a React board on localhost:5173 that ships an offline FIXTURE demo mode. The bundle embeds the skill only; product source stays in the repo.

Measures one university department's research output: matches author affiliation strings against a college's known name variants (detecting same-named departments at other universities), backfills papers that only name the university by faculty author IDs, classifies each paper by journal tier (CAS partition with Top flag, JCR quartile, CCF, Chinese core via a CNKI export merger), splits staff from student first authors, and writes a workbook with a dashboard sheet, a filterable paper list, a teacher-by-direction matrix and per-teacher statistics. Student authorship is inferred from affiliation and staff lists, not verified against enrolment records. Requires Python 3.9+ with pandas and openpyxl (pip install -r requirements.txt); papers come from the OpenAlex API, whose free per-IP quota is limited, and Chinese-journal coverage needs a manual CNKI export.

Two-way sync between a local LaTeX folder and an Overleaf project over the Overleaf Socket.IO + OT channel instead of a Git bridge: pull with automatic git init, dry-run-first push, three-way diff, background watch, and conflict copies that never overwrite.

Read PLC projects inside DSH — decode CoDeSys .project and Delta .dvp (both need Python 3.9+ on PATH) and Siemens S7-200 SMART .smart/.smartlib into readable reports and structured data, plus project diff, address audits, element-level read/write analysis and ladder-logic reconstruction.

After an agent's own bash or shell command, re-observes the files that session had observed, so FS_STALE_VERSION no longer fires on self-inflicted writes but still fires on writes by others.

Bilibili companion for DSH (MIT fork of Alterssss/dsh-bili-whale; do not install alongside the upstream — both register the same `/dsh-bili/` settings routes and share `~/.dsh/bili-whale.config.json`): wakes on a schedule, reads videos through subtitles, danmaku, hot comments and covers, writes two diaries, and shares at most one pick per wake by commenting on the video and @-mentioning the owner (daily quota 3; DM when `share.via=dm`). Inbound DMs, commenter names, summary requests and live danmaku are normalized and scanned for injected instructions before they reach the model; only the owner DM channel can carry an authorization prefix.

Local-first, file-native long-term memory for DSH with per-turn preflight, compact context injection, citations, human-approved writes, manual consolidation, nightly report-only maintenance, and a Web panel that is read-only for knowledge while allowing workspace-config edits.

Copies the compaction checkpoint into the sticky-notes archive when context is compacted and wakes the session to write what was lost.

Write each session's summary as a Markdown file in its workspace, with a page to browse, regenerate and delete them.

Persona slot for DeepSeek Harness: each session holds exactly one persona that a later write replaces, a global persona covers every session without its own, a /persona command edits both in conversation, and the Installed page exposes a multi-line panel for the global persona. Persona length is unbounded. Supports DSH 0.2.1-rc.2 through rc.5.

Local Ollama worker tools for DeepSeek Harness: model inventory, a single unverified ask, and a batch runner that chunks long text, checks every item deterministically, returns a small verbatim sample for review and can optionally write the labelled result inside the session workspace.

Manages the lifecycle of the local UM GenAI proxy (127.0.0.1:8788) that DSH's um-genai provider talks to: wakes it on demand for llm/stream calls, stops it after an idle period, and reports health plus a read-only status endpoint. Also ships a DSH settings page for port, always-on mode, idle minutes, interpreter and paths, plus the UM account and proxy API key, and a button that writes the providers.um-genai entry into the profile patch and fills in its model list, since a fresh profile has no provider and the model picker would otherwise stay empty. A renewal watchdog logs in to the campus gateway on a hidden Windows desktop when the upstream token expires. The proxy itself is um_opencode_setup.py, a single-file tool distributed by the University of Macau; a copy is vendored in vendor/ so a fresh install works out of the box, while a copy in ~/.um-opencode/ takes precedence when present. The plugin still does not produce tokens: those come from logging in to the campus gateway. Requires Windows (process discovery and idle shutdown use netstat/taskkill; renewal uses PowerShell and a hidden desktop), Python 3.10+ and a University of Macau account.

Signs in to LightVela by WeChat QR code, reads the model gateway key from your own instance and writes the gateway into llm-pi-ai so its models appear in the picker.

WYSIWYG Markdown notes editor in the DSH right sidebar, with tabbed documents, table editing, chat integration, assistant read/write tools, and local file export.

COROS Sport panel for DeepSeek Harness: sign in to COROS inside the panel, read the account, bound devices, recovery, fitness, recent workouts, daily health and the training schedule in one dashboard, and get a plan from the model with /coros-plan. The plan page writes a fixed four-week, 16-session plan to the account only after a two-step confirmation.

Replaces the composer Effort switch with a Claude Desktop-style five-level slider (Low/Medium/High/Max/Ultracode): a pill opens a popover above the composer, dragging snaps with a spring and writes once, and Ultracode maps to the model's maximum reasoning effort. Ships the Ultracode WebGL2 energy grid, a per-session Ultracode mode, and passes modifier-click through to the shipped level list.

Conserve le préréglage de permissions Auto après redémarrage de DSH : la migration de déchargement auto-review à l'arrêt ne transforme plus une Auto Session active en accès Full, et une Auto Session enregistrée peut s'ouvrir même si son integration est absente.

Hook pré-exécution qui refuse uniquement les écritures dans une zone de résultats en lecture seule de racines protégées. Tous les autres chemins passent.

Ajoute une puce Git près du composeur pour ouvrir un atelier central (changements, historique, fichiers, tags, stash) et une section Git dans les paramètres. OAuth device flow multi-comptes et PAT, avec permissions par dépôt (full/read/none). Un commutateur global contrôle les écritures ; les lectures sont limitées à une liste de répertoires autorisés. Refuse le push si aucun compte ne peut écrire sur le remote cible. Diffs affichés sur place avec numéros de ligne, 4 mises en page, emphase par mot, coloration syntaxique et recherche Ctrl/⌘+F ; rendu inline de Markdown et d’images, recherche et génération de messages de commit depuis le diff staged via le LLM de l’hôte. 8 outils Agent, 51 endpoints HTTP locaux, 8 langues ; tokens chiffrés au repos par AES-256-GCM.

Préréglage de workflow de livraison pour DeepSeek Harness : protocole d'agent et 7 outils qui conservent les demandes utilisateur mot pour mot, enregistrent les décisions de retour, planifient avec critères d'acceptation, consignent les vraies commandes de vérification et codes de sortie, et écrivent les résumés de fin dans des fichiers JSONL par projet.

Boucle de revue du travail pour un coffre de type Obsidian : rédige le Dream du jour et un plan T/G numéroté, n'exécute que les éléments marqués routine (lecture seule par défaut ; commits locaux uniquement avec --live), regroupe sept jours en routine de semaine ISO et affiche chaque jour/semaine dans un panneau latéral en lecture seule.