Plugins

Integrates TypeSafe Jev for skill and file selection, task supervision, shared-finding corrections, tool-output filtering, approval assistance, and recorded-stage navigation in DeepSeek Harness; all features are disabled by default.

Causal memory for dsh: a zero-dependency MCP stdio bridge that mounts the 17 tools of the causal-memory server under clean names (no mcp__ prefix) — decision → outcome edges with typed relations (caused / enabled / prevented) in a local SQLite store that survives compaction — plus a pinned system-prompt section. Requires the causal-memory server binary on PATH (pip install causal-memory, or a prebuilt release).

Multi-model council for DeepSeek Harness: independent answers, anonymous peer reviews, and an inspectable final decision.

Right-side details panel that turns the Matt Pocock skills collection into a DSH quest board: wayfinder decision maps, per-issue triage / fix / discuss / execute actions, 9-item core-skill probe plus 20-skill radar, environment checks, and one-click handoff across sessions.

Classifies pending tool calls with Jev (TypeSafe AI's non-generative decision model) on tools/pre-execute — confident high-risk calls are denied, ambiguous ones escalated to approval — and auto-approves clearly-granted reversible calls on approval/request behind argument-evidence gating. Also subclasses the session-reference resolver so snapshots keep Jev-scored messages instead of dropping oldest-first. Degrades to stock behavior on any provider failure; shadow mode with a /jev-stats command; TypeSafe or OpenRouter endpoints; 64 tests.

Jev DSH 决策引擎｜面向 Agent Harness 的结构化决策插件。原生支持 DeepSeek Harness，通过 iPolloWork 支持 OpenCode、Codex Harness。

Jev-powered decision layer plugin for DeepSeek Harness

StrataGate gives DeepSeek Harness evolving, local-first long-term memory for user preferences, project decisions, conversations, and tool results across sessions; it layers history with user-tunable time decay, reinforces memories through real answer use, shows workspace-aware usage counts, and keeps recalled evidence traceable to its original messages.

A mid-run steer input for DeepSeek Harness Web: while the agent works, an input bar appears above the composer to inject modification requests into the next decision via agent.steer(), without stopping the conversation. Drafts persist per session, unread injected messages can be recalled from the bubble, and an optional guiding prefix is supported.

Cross-session memory for DeepSeek Harness: agents save, search and update user preferences, facts and project decisions; important entries are injected into later sessions, and a Settings section lists, edits, archives or deletes them.

AI Company Framework: turn one sentence into a multi-agent company — 15 flat Skills (boss/pipeline/role-template/11 roles/generic customer-memory), 7 workflow templates, AgentTeams runtime + activity panel, employee sidebar with per-session isolation, Feishu bot bridge (boss and per-employee bots, group routing, autonomous push, human decision gate, external-contact guide), per-customer SQLite memory with three-layer memory (cross-group merge), /company mode switch; no RAG.

Runtime requirement drift guard that keeps long-running DSH agents aligned with the user's approved goal, constraints, and decisions.

Laya typed decisions (`noul` yes/no, `choice`, `score`) as a first-class Cordis service (`ctx.laya`) plus the model-visible `laya_ask` and `laya_plan` tools. The plugin itself installs and downloads nothing: it is a client of a `laya-mcp serve` sidecar you start yourself, and it reports that the sidecar is not running instead of failing obscurely at the first tool call.

Per-step model router: named presets with ordered image/keyword-group rules route each agent step across every mounted provider and model, adding collaboration flows for image transcription, an image guard, a multi-plan quota panel, and per-step decision observability.

Notification sounds for DeepSeek Harness — a soft chime when the agent needs your decision or finishes a task, with a Settings section styled like DSH's own · 需要决策或任务完成时播放提示音，点击系统通知可回到窗口，设置面板可调音量与音效

AX-first macOS computer use for DeepSeek Harness: accessibility snapshots with stable refs, background input, Unicode typing, menus, apps, set-of-mark screenshots, on-device OCR, verified actions, and optional Jev (TypeSafe System One) semantic guards — 1

AX-first macOS computer use for DeepSeek Harness: accessibility snapshots with stable refs, background input, Unicode typing, menus and apps, set-of-mark screenshots, on-device OCR and verified actions — 11 tools bridged to the macos-cu CLI.

Website cloning methodology skill: a real-source-first decision tree covering static sites, React/Vue/Next content sites and WebGL/Canvas-heavy fronts, mandatory verification of executable code quoted from second-hand AI analyses, and 13 zero-dependency Node scripts for CDP browser recon, asset/font harvesting, route crawling, sourcemap hunting and pixel-diff auditing.

Dispatch subagents with named profiles (model, reasoning effort, and tool scope per task), safety checks, cost estimation with savings comparison, and a full dispatch decision ledger.

DSH Desktop enhancement plugin: it puts verifiable facts on the table. Constraints are injected only when relevant, so idle chat costs no extra tokens; injection is layered (session-stable text in the system prompt, volatile content as a tail snapshot) and the prompt cache stays valid. Discipline: each turn is classified (question / research / discussion / diagnosis / execution) with boundaries enforced - no file edits on question turns, no unsolicited fixes on diagnosis turns; citing code lines not opened this session, negative claims about unseen symbols, requirements that do not match the deliverable, and dressing up “I did not check” as “waiting for your decision” are all called out; replies are written for humans - a code, field name or id must be explained in plain words the first time it appears. Memory and knowledge: 75%/90% context warnings plus an exported session recap (say "continue" in a new session); project knowledge accumulated across sessions by working directory, auto-captured and deletable by id. Inspectable artifacts: task contract / change record (auto-logged, unverified items flagged at delivery) / hypothesis record (conclusions must pass an evidence gate) / design decisions. Dashboard and personas: routing, trigger hits and block assembly are written to lume-metrics.jsonl and queryable via lume_metrics; personas are distilled from chat logs, novels, scripts or character sheets, with long-term and temporary memory, a memory star map and persona card export/import.

Proactive multi-model scheduling plugin: multi-source signal intake with dedup and urgency ranking, a four-way execute/defer/dismiss/ask-user decision engine, DAG plan generation with parallel model execution, quality reflection with automatic retry and model switch, long-term memory for task patterns and lessons, and strategy evolution via genetic algorithms with sandboxed evaluation and canary rollout.

Community search, similarity analysis, and install/dedupe/build-it-yourself decisions for the dsh-plugin ecosystem.

Heads-up Texas Hold'em training with an independent AI opponent, persistent public-action memory, and decision reviews.

TypeSafe Jev (System One decision model) as the jev_decide tool for DeepSeek Harness: noul, choice and score questions return calibrated probabilities; credentials reuse the DSH credential seam; 429/529 backoff retry built in.