Plugins

Image-to-text input for the Web UI: paste or drag an image and it is transcribed into structured text and sent, giving text-only LLMs image-input takeover (OpenAI-compatible vision API).

Headless-browser verification tools so the agent opens the page it just built, reads the rendered DOM and computed styles, checks the console, and screenshots the result; ships a frontend-verify skill.

Browser operation by accessibility refs: navigate, snapshot the page as a ref tree, act on elements, and wait for conditions; the origin policy follows the session approval stance and is enforced by request interception.

Real-browser automation for dsh via an MCP server (Playwright core): navigate, click, fill, type, screenshot, and act on accessibility snapshots with numbered references - all 22 web_* tools re-exposed to the harness over stdio.

Render Markdown + JSON into self-contained HTML documents, cards, dashboards, and galleries (CLI + artifact_render tool).

Human-in-the-loop tools (`human_task` / `human_task_ready_check`): pause the agent until the user performs a real-world action or returns an observation (GUI operation, game testing, visual verification, hardware checks), gated by session consent and an AFK presence check, returning a structured JSON result via a web task dialog.

Observational file-mutation rollback for the fs tools: write/edit pre-images are checkpointed as git blobs or snapshots and restored through a model-facing rollback_files tool or a /rollback command.

Ecommerce operator workbench: CSV batch preview with marketplace column adapters, reproducible profit/six-dimension scoring, public product-page snapshots, Chinese skills, and a replaceable store-policy KB.

Expands #tag shorthand into snippet-library text before messages are sent; multi-library, aliases, variables and recursion guards.

Doctrove: versioned library documentation retrieval MCP server for coding agents — catalog lookup, version selection and focus-ranked doc extraction, with zero runtime dependencies and a dsh bundle.

Desktop (native) control for dsh as a stdio JSON-RPC bridge over a Python core: capture the screen, inject pointer/keyboard input, drive accessibility-tree actions, and gate high-risk operations behind required confirmations.

Scans a project's declared build commands — Makefile targets, package.json scripts, just recipes, deno tasks — and exposes them as ready-to-run agent tools.

Slash-command console for dsh: /web start|stop|restart|status, /tunnel SSH forward management, one-shot /ask, and a console TUI launcher.

Turns markdown into a WeChat Official Account (微信公众号) draft: inlines every CSS rule onto its element because the editor discards `<style>` blocks and class names, uploads each local image through WeChat's own endpoint so it is not filtered out, and creates the draft; it never broadcasts.

Connects DeepSeek Harness to AsimovBox for creating, updating, rendering, and finishing videos through API-key-authenticated tools.

Sync the installed DeepSeek Harness plugin manifest to a GitHub Gist (OAuth login, one-click export/import with automatic dependency install).

Local-life toolkit for DSH agents: budget planning, bill splitting, unit-price and discount math, common unit conversions (jin/li/chi, C/F, L/gallon), and trip checklists.

Education toolkit for DSH agents: lesson-plan skeletons by grade band, quiz scaffolding and answer-key validation, analytic rubrics, Anki/markdown flashcards, and readability levels.

Auto-supplement QMD semantic search when the model greps/globs a knowledge-base directory: zero-dependency DSH plugin, async injection via next-step, no manual invocation.

Project memory for coding agents on DeepSeek Harness: similar past edits surface as precedents, out-of-scope edits are blocked, and cognition persists across sessions (47 upstream-verified fixes, gold-free 20-task run).

PLUR memory rendered into the system prompt on each assembly rather than exposed as a tool call, so the block is replaced instead of appended and context stays flat across a session. Fully local hybrid search (BM25 + BGE fused with RRF), plain YAML storage you can edit, per-workspace scoping, and a /plur-memory browser.

LLM-driven remember/forget memory: after each turn a model decides what to keep (MemoryDirector), injects relevant memories before each step, dedups, and persists cross-session.

Compaction backend replacing LLM summarization with a deterministic semantic extractor (keeps code/paths/commands, drops chatter) plus 28.4x KV-compression accounting.

Session context triage for long DeepSeek Harness sessions: deterministic screeners trim stale, duplicate, failed, and oversized content through the official ctx.compaction seam.