Plugins

Guarda de consistencia entre documentación y código: audita encabezados de versión, tablas de cambios, árboles de directorios, recuentos de módulos y pruebas y referencias cruzadas entre documentos en busca de desviaciones, con sugerencias de corrección de solo lectura ordenadas por severidad.

Guarda de arranque que repara registros de sesión corruptos, preinspecciona la composición de plugins, verifica con vm los artefactos de cliente, hace pruebas de humo de apply() en el host y pone en cuarentena los paquetes que provocan cierres, para que un plugin roto nunca inutilice el arranque.

Protocolo de colaboración multiagente independiente del proyecto para DeepSeek Harness: guardas de bucle que detienen bucles descontrolados, un formato de traspaso de seis campos, enrutado de revisión por niveles de riesgo y una secuencia fija verify→commit→push→review→close. Una skill en lugar de las mismas reglas copiadas en las instrucciones de cada agente.

Paquete de protección contra mojibake para DSH: detección de texto ilegible basada en características antes de escribir con revisión y liberación (un reintento idéntico pasa), más auditoría de codificación de pwsh antes de la ejecución (ANSI/GBK/UTF-16/Default). Windows y chino primero.

Protege la caché persistida de proyección de sesión: degradación JSON por fila para que una unidad de proyección defectuosa (p. ej. un plugin de terceros que guarda un Map) no pueda volver a bloquear los títulos de sesión ni toda la caché, más una autorreparación al arrancar que rellena los títulos faltantes.

Plugin de protección de escritura consciente de Git para DeepSeek Harness: bloquea las escrituras de archivo completo que sobrescribirían cambios sin confirmar del usuario, e informa al final del turno de qué ocurrió con los archivos protegidos

Añade comprobaciones de políticas para operaciones peligrosas, redacción de salida y un flujo de trabajo de revisión de seguridad.

Analiza skills y configuraciones MCP en busca de inyección de prompts, homóglifos, Unicode oculto, comandos de shell peligrosos y fugas de credenciales.

Control de contenedores tipado y protegido: ps/logs/inspect/exec/start/stop y compose up/down con salida JSON, selección consciente del proyecto y operaciones destructivas sujetas a aprobación.

Harness-neutral Traditional Chinese enforcer + offline converter (skill, DSH bundle, CLI). One script axis either way + Cantonese/Japanese-only filters. Converts both ways; optional wording preference. Also: an LLM output guard, cp950/GBK console scan.

A structured decision layer for the DeepSeek Harness agent loop: a danger-call gate, output self-check, tool narrowing, and loop guard backed by a pluggable adjudication model.

LAN access for the official DSH Web UI through a gated reverse proxy: DSH keeps its own loopback binding, and a second port adds a password gate, self-signed HTTPS and per-device pairing that can be revoked individually, with a QR code to open it from a phone.

Host- and agent-layer DeepSeek Harness plugin that prices every automatic context rewrite before it lands, asks the human by default, and reports what the provider had to re-read cold.

GitHub operations for DeepSeek Harness: releases, tags, a guarded generic API pass-through, workflow runs, secrets and sanitized mirror publication.

Pre-execution safety valve for DSH that judges with TypeSafe Jev: it hooks tools/pre-execute and decides every bash/pwsh call by first applying offline static rules, then by asking the Jev model — the System One model from TypeSafe, which returns a structured decision instead of prose — one yes-no question ("will this command irreversibly delete or overwrite real data?"), turning the answer into allow, revise, block or escalate; revise hands the model a safer rewrite, escalate offers a one-shot human token; exhausted credit or a missing API key degrades loudly instead of failing silent, and a missing key is requested in the conversation itself (`guard key set`, read from stdin); every verdict is appended to a shared audit log and the text people read is bilingual zh-CN/en.

Customize the dsh UI theme: 12 accent presets, dark-mode aware accents with a contrast guard, density, font family, animation toggle, and raw design-token overrides.

Zero-intervention permission gate for DSH agents on macOS: auto-allows local writes made reversible by git or Time Machine, blocks network, package installs, process control and sensitive-path reads, with full audit logs.

Agent 行为守护桌宠：跨 agent 监控可疑外传（加密打包直传对象存储 / DNS 外泄线索 / 敏感密钥被读），发现即告警，明确确认后按「一次一个」终止目标 agent；事件写入只追加 guard-events.jsonl，并通过 DSH host 工具 + MCP(stdio) 供其它 agent 查询

Consistency guard for long-form writing: scans conversation prose for character, timeline and numeric setting conflicts across chapters, then jumps straight to the conflicting message.

Auto-archives sessions idle beyond a configurable threshold, with safety guards (never touches live or running sessions), plus a full archived-session management UI: list, unarchive, safe delete and content preview.

Keeps a redundant sandbox-escalation argument from failing a tool call. Escalating tools (pwsh, bash, write, edit) advertise the full sandbox_permissions enum, but DSH only accepts a level strictly wider than the one already in effect — a check its own source calls 'deliberately not a schema constraint'. A model that reflexively passes the argument therefore picks the level it is already at and the call dies before running with 'sandbox escalation to "workspace-write" is not strictly wider than this call's current "workspace-write" mode', costing some models a whole turn of retries. This plugin registers one tools/execute waterfall listener and, only on that documented rejection and only when the arguments really carried an escalation field, re-issues the identical call once without it. Safe by DSH's own documentation: the rejection precedes any execution ('nothing has run'), and the corrected call cannot match again, so the retry is loop-free. Reproduced and verified end to end — before: isError true, the command never ran; after: the command's real output, isError false, one tool/call and one tool/result. Zero dependencies.

Control panel for a local credential-redaction gateway in DeepSeek Harness: gateway and model health, audit metrics, a dry-run leak tester, an exemption whitelist, encrypted key management, credential-protected model entries and deployment controls.

One install for the community's verified DeepSeek Harness fixes: a bundle that mounts every guard and patch in a single profile line, plus a read-only projection-cache footprint report.

DSH 插件：移动端切换会话时抑制 composer 的自动聚焦，避免软键盘被顶起