jev 27

TypeSafe Jev as a first-class Cordis service and three model-visible tools for DeepSeek Harness — offline by default, egress disclosed, never default-on.

Millisecond System-1 judgement for every tool call in DeepSeek Harness: Jev-powered risk classification and evidence-gated auto-approval, fail-closed by construction.

Gives DeepSeek Harness three automatic Jev judgments — prune oversized tool output, screen fetched pages for injected instructions, suggest which skill to use — plus the jev_ask and jev_gate tools, a persistent judgment ledger, and a /jev-status diagnosis

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

DSH 插件：把 TypeSafe Jev（System One 决策模型）注册为 agent 工具 jev_decide。调用 https://api.typesafe.ai/v1/systemone，让 agent 在路由/分级/guardrail 等场景拿到带校准概率的结构化判定（noul/choice/score），而不是让大模型瞎猜。

TypeSafe Jev (System One decision model) for DeepSeek Harness: typed decisions in three question primitives, confidence-gated routing, a cost meter, and an automatic prompt-injection guard over untrusted tool results.

Jev (TypeSafe System One decision model) for DeepSeek Harness — seven tools: jev_decision (up to 25 typed questions in one parallel call, measured median 266-484 ms), jev_choose (rank 2-10 approaches), jev_batch (one set of up to 5 judgments over 1-20 tex

把 TypeSafe / Jev（System One 判断模型）的构建技能注册进 DeepSeek Harness 的全局技能层：分类、路由、打分、抽取、置信度门控这类「让 AI 做判断」的任务，每个会话都能直接调用。不含任何密钥，API key 由使用者通过 OPENROUTER_API_KEY 环境变量提供。

Jev-scored context compaction for DeepSeek Harness -- keep the messages that matter, compact the ones that do not

Per-turn model and reasoning-effort routing for DeepSeek Harness -- one Jev choice question per turn

DeepSeek Harness plugin that wraps @stardeckai/lgtm — a Jev-powered linter for tests that pass but prove nothing. Ships the CLI as a dependency, reuses the TypeSafe credential stored in DSH, and exposes lgtm_audit / lgtm_status to the agent.

Fact-checking skill for DeepSeek Harness: verifies claims against the open web instead of model memory, cross-checks independent sources across domestic and international search ecosystems, and can route three factual judgements (does the evidence support

Plan Checkup for DeepSeek Harness: checks each step of a plan with Jev or a local LLM before you approve it, and marks the results on the plan review card.

Cut LLM costs: route routine tasks to cheap subagent models (DeepSeek Harness plugin)

TypeSafe (System One / Jev) decision primitives as a host service: typed noul/choice/score judgements through one named-question registry, with credential-reference key resolution, bounded timeout, retry, circuit breaker, cache and decision logs. Other pl

Jev (TypeSafe System One) adapter for the dsh decision layer

Jev (TypeSafe System One) as a Cordis service for DeepSeek Harness: one headless decide() call, no tools, no policy.

DeepSeek Harness plugin: Jev-backed guardrails for prompts, tool calls, tool results, and model responses.

Pre-execution safety valve for DeepSeek Harness: a TypeSafe Jev question decides whether a shell command/script would irreversibly delete or overwrite real data. Four states (allow/revise/block/escalate), mounted on tools/pre-execute. Zero dependencies; W

Semantic find tool for the pi coding agent and the DeepSeek Harness — a cascade-search port of jegrep / oh-my-pi jfind

dsh plugin: Jev-guided dispatch advice — before a turn starts, a TypeSafe Jev call decides whether the task belongs to a registered agent and which one, then injects a four-line, ignore-if-unsuited recommendation. Jev-optional: no key means silence.

Jev (TypeSafe System One) gives the human a second opinion, not the agent. When DSH's model asks you a multiple-choice question, dsh-jev-advisor builds a structured Jev request and floats the recommendation, probability spread and confidence beside the op

dsh Web-only plugin: jev System One lowers reasoning effort to low for simple turns, with a dismissible composer hint

Jev/System-One semantic pruning of stale tool results for DeepSeek Harness; keeps conversation text verbatim and forgets stale tool output before summary compaction is needed

Codex-style auto-approval for DeepSeek Harness. An independent reviewer decides allow/deny, fail-closed, with a rationale for every decision. Engine: an LLM route, or TypeSafe's Jev (`reviewer.engine`). · DSH 插件：独立复核器裁决，引擎可选 LLM 路由或 Jev。

Jev (TypeSafe System One) judgments at the DeepSeek Harness agent-loop gates: pre-step, pre-execute, post-execute, and turn-stopping. No UI dependencies; mount it in any Harness composition.

The /JevLoop control panel for moqi: gates and API key from dsh-jev-loop, rendered through moqi's tuiHost seam.