Plugins

Vision-Brücke für reine Textmodelle: Bild einfügen, strukturierte JSON-Belege erhalten (OCR, Layout, Semantik).

Software-engineering method pack for coding agents, with skills for baseline-first planning, systematic debugging, prompt hygiene, verification before completion, and repair/retirement tracking.

Websuche-Brücke für reine Textagenten: das Web oder X befragen, strukturierte JSON-Belege erhalten (Suche, Abruf, Zitate).

Cleans AI traces from Russian text: finds chatbot copy-paste artefacts (ChatGPT, Gemini, Grok, Perplexity, DeepSeek) and rewrites the text into natural prose on request; 39 regex markers with an evidence registry, offline, text-only bundle.

Deep-read books and articles in five modes (quick, deep, knowledge map, Feynman, full book) with claim-evidence-data reports, Mermaid/XMind mindmaps, batch comparison, budget preflight, transparent background-job progress, and MD/HTML exports from URLs, files, or pasted text.

Academic writing guard (EN/ZH bilingual): removes AI-style defensive writing, protects scientific evidence (numbers, citations, claim strength, null findings, scope) during polishing, and calibrates writing toward a target journal (Journal Profile + Journal Fit). Deterministic local regex/statistics, zero network/LLM; writing_rules / writing_audit / writing_style_profile / writing_journal_profile with auto-audit on paper writes.

Free vision bridge and image generation for text-only models: paste-image reading, GLM-4V-Flash and Gemini engine failover, ModLens-style structured evidence, and a seeded free vision model route.

Vision for text-only DeepSeek via Doubao Web by default (zero-cost, no API key — drives your logged-in Chrome through a Windows CDP bridge), with Antigravity IDE quota (flash/pro) or Gemini fallback; auto detail escalation, vision evidence memory with compaction rehydration, content-hash cache, and a bilingual client panel.

Suite von Recherche-Agenten: langfristige Zielsteuerung mit auditierbaren Belegketten, Cron-Scheduling, Multi-Agent-Expertenteams, sich selbst weiterentwickelndem Recherche-Gedächtnis (FTS5 + RRF-Recall), Projekt-Workspaces und einer eigenen Workspace-UI.

Fügt DSH Web einen eingebauten Plugin-Katalog hinzu, mit Suche, Statuserklärungen und sicheren Schaltern für geprüfte UI-Plugins.

Let agents search, read, and cite your local Zotero library: find papers, view notes and annotations, pull evidence by question, open the source document, and generate citations.

Technical-preview local graph memory for DSH and Codex: recalls owner-global and current-project evidence before each turn, writes USER and ASSISTANT records separately, and preserves project, session, actor, and source provenance for a visual atlas and cited knowledge pages.

Always-on dependency security for DSH plugins: tracks exact installed paths, OSV vulnerabilities, npm releases, and breaking-change signals, then routes project evidence to a DSH Agent.

Hängt lokale, hashverkettete JSONL-Quittungen für Tool-Ergebnisse und Sitzungsereignisse an, ohne Prompts, Tool-Argumente, Ergebnistext oder rohe Sitzungs-IDs zu speichern.

Gehosteter, kostenloser Vision-Sidecar für DeepSeek Harness mit dauerhaften Sitzungsbelegen.

Wächter für Engineering-Disziplin: hinterfragt die Anforderungen vor der ersten Änderung, erzwingt Rot/Grün-Testnachweis-Gates und prüft die Lieferung mit einem abgespaltenen Gegenspieler (grill-requirements-Skill + Tool-Policy-Gates).

Auditable vision and cross-platform Computer Use runtime for DeepSeek Harness with source-preserving evidence.

Evidence-first guard: warns the agent when it claims success without a matching tool-execution record in recent context.

Evidence-bound negative-knowledge ledger: persists disproven paths (command_failed, file_missing) with their outcome and precondition evidence, then warns or blocks repeat attempts until the evidence changes.

Evidenzbasiertes Betriebskit: sechs rein lesende Tools (Fähigkeitskatalog, stufenweise Workflow-Pläne, Reader für gepackte Skills, begrenzte lokale Gedächtnissuche, Release-Audit für Repositories, Release-Checkliste) sowie fünf gepackte Skills — plant und prüft, schreibt nie remote.

Evidence-first self-evolution controller: bounded Cordis candidate generation, one-shot real-Loader admission, Harbor evaluation, and a crash-resumable journaled lineage.

Adds persistent execution contracts, recursive work graphs, critical clarification, and evidence gates for long or underspecified Harness tasks.

Datenschutzfreundlicher lokaler Scanner für Sitzungs-/Workflow-Metadaten von Codex, Claude Code, WorkBuddy und CodeBuddy; speichert aggregierte Belege und empfiehlt integrierte DSH-Agent-Presets, ohne Inhalte zu behalten, ein Netzwerk zu nutzen oder Presets zu verändern.

Falsification-driven correction loop for agent sessions: risky work commits an evidence-checkable claim first, deterministic gates verify it, falsified claims force mutually exclusive replacement hypotheses with discriminating experiments, and every event lands in an append-only evidence ledger.