Plugins

Fronts stdio MCP servers with three router tools (list_capabilities, describe_tool, call_tool) so schemas load on demand instead of riding every request; shapes every tool result through a deterministic output ladder that archives the original first (retrieve_spill reads it back); prices tool groups in tokens and can hide built-in groups per agent. Bridges to a standard-library Python router, so Python 3.9+ is required.

Compaction fallback for compaction-basic: when its LLM summary throws, comes back empty or cannot fit the window, the span is compiled deterministically with a seq pointer on every line (vendored dsh-compaction-instant compiler). Adds recall and search tools, /recall and /context commands, an idle trigger at 45% of the window and an on-disk span archive. The engine mounts as one row in the agent preset (docs/dsh-integration.md).

One settings page that lists every subscription sign-in the harness can offer and walks you through it — ChatGPT (Codex), Claude, GitHub Copilot, OpenRouter, Kimi and xAI — reading the list from the authorization seam instead of hardcoding providers, and showing credential records no registered flow claims any more.

Pool AI provider accounts per platform and route them through a local loopback proxy that bridges the OpenAI, OpenAI-Responses, Anthropic and Gemini protocols, then write the proxy into Codex, Claude Code, Gemini CLI, Grok, OpenCode, OpenClaw and Hermes config files with snapshot-and-rollback safe writes; registers no agent tools.

Publishes a lesson into the DSH conversation: the dependency spine, the question and the vault's drawings, drawn in the vault's own palette.

Ren'Py galgame production workbench: a host-side project service (write gateway plus git snapshots, a dialogue/scene dialect parser, derived progress, human-only review stamps, pinned-SDK validation and playtest, local publish) behind one seam, driven by an embedded web panel and agent tools.

Registers the imagegen skill on the skill registry, so generating and editing raster images needs no hand-copied files under $DSH_HOME/skills. The agent drives a bundled image_gen.py CLI against any OpenAI-compatible GPT Image endpoint for text-to-image, reference-image editing, and transparent-background cutouts.

WeChat Official Account channel for DSH — WeChat messages run agent turns and the answers are pushed back through the customer-service message API, with signature-checked callbacks and an openid allowlist.

Cron job board: schedules local scripts or dsh session tasks (fresh or pinned sessions) by cron expression, with centralized environment-variable management, run logs, and concurrency and timeout control.

English word bank for one or more children: words marked on textbook or workbook photos (highlighter, red pen) are found by their marks, then translated and stored in a local SQLite bank with an occurrence count; the clipboard picker, the photo route and dictation all land in the same bank. High-frequency words surface by count, three correct answers in a row flag a word as learned, and any selection prints as a memory card set (a word split into chunks, real IPA per chunk, a mnemonic hook, an A4 sheet of eight) in HTML, PDF or Word, with a print-and-answer exam page whose four options are set against each other for meaning. Requires DSH 0.1.7 or newer (1.0.3 moves to the 0.1.7 settings contract); on 0.1.5 the settings panel does not appear - install dsh-word-vault@1.0.2 instead.

Layered memory for DSH whose agent-facing surface is 10 tools: high-privilege writes (memory_add, memory_delete, memory_import), rumination controls (memory_ruminate, _cancel, _status), and a memory graph (memory_search_graph, memory_expand_graph_node), built over an L0-L3 distillation pipeline that recalls and injects relevant memories before each model step.

Same-origin in mechanism with the two loudest lines in context compression. The code-skeleton gate follows the skeletonization approach of Headroom (Apache-2.0), whose published headline is 20% fewer tokens for coding agents and 60–95% fewer tokens for JSON, same answers. The estimator channel follows TokenPilot (arXiv:2606.17016), which reports up to 60% lower cost for long-session agents. Both figures are theirs, quoted as-is; this plugin ships no benchmark of its own and claims no reduction of its own. What it adds for DeepSeek Harness: choose a compression profile, set the Auto Compact trigger level and toggle code-skeleton compression from one settings section, with exact DeepSeek V4 tokenizer measurement, same-revision count verification, and fail-open behaviour that keeps the original tool results on unsupported models.

Session-affinity routing for LLM gateways: injects the session ID into every request sent to NewApi-style gateways so the gateway can pin requests per session and raise prompt-cache hits; takes over the official pi-ai routing on install and restores it on uninstall.

Multi-platform LLM balance and quota panel: configure DeepSeek, OpenRouter, Kimi, Zhipu GLM, MiniMax, NewApi or custom endpoints, poll on a schedule, draw balance and quota trend charts, and push threshold and window-reset notices through multiple channels.

Pixel-font "deep seeking" thinking status with an animated DeepSeek blue whale (20 spout frames and 32 tail-wag frames).

Literature search for DeepSeek Harness over the official PubMed NCBI E-utilities API (esearch/esummary/efetch/elink) and Google Scholar (SerpApi engine, or direct HTML parsing because Google publishes no Scholar API). Zero runtime dependencies.

Six independently togglable modules in one DSH plugin: per-turn CNY cost, DeepSeek account balance, PDF/Word/Excel/CSV read-write tools with right-sidebar preview, a compiler/runtime environment registry, and run/build tools that execute a program or an MSBuild project by environment id. Toolchain paths are resolved at run time (Visual Studio install-root scan, %ProgramFiles%, PATH) and can be overridden per environment through plugin config, so nothing is pinned to the author's machine.

Local-first voice conversations for DSH, with local speech recognition and synthesis and optional external providers.

Memory Fortress for DeepSeek Harness: 5-track layered memory, self-evolution, knowledge graph (SQLite+FTS5), status pill. 本地记忆堡垒：五轨记忆·自我进化·知识图谱·状态灯。

DeepSeek Harness web-search provider backed by a SearXNG instance: metasearch with no vendor API key and no account tying the queries together.

Cross-session long-term memory for the DeepSeek Harness: a two-phase extraction/consolidation pipeline over a file-based memory workspace, recall injection, a citation feedback loop, lease-based job semantics, plus a settings page, a sidebar status panel

Canonical project memory for DeepSeek Harness: one governing article per asset, an append-only journal beneath it.

Agent-first task kanban for DeepSeek Harness web GUI, modeled on codeg-plus's 任务看板: four codeg-plus columns (todo / inProgress / attention / done), human-only acceptance, workspace-bound claims, comments and a live SSE board view. 参考 codeg-plus 任务看板语义的 ds

Out-of-tree dsh plugin: GitLab, GitHub, Gitee, Gitea, and Bitbucket tokens stay out of the model context, stored encrypted (AES-256-GCM) in a plugin-owned file; the model calls forge API tools on demand, and the web settings page manages sites and tokens.