verbs

Pull a decorrelated second opinion from a DIFFERENT model on a LOAD-BEARING judgment: a design fork, a pre-commit plan, an irreversible or outward-facing call. Reversible calls just decide. `--panel` = blind critics on a prepared plan. NOT code-diff review (review), NOT mechanical work for Codex (handover), NOT sharpening a fuzzy idea (grill).

Build or fix UI surfaces with a committed point of view: pages, components, layout, typography, motion feel, design review. Triggers: 設計, 做頁面, 做組件, 不好看, 很醜, 突兀, 不協調, 字體, 排版, 樣式, 前端, UI, 截圖, 很死, 死板, 生硬, 沒動效, 沒回饋感, 切換好硬, design, build a page/component, make it look good, typography, feels dead, stiff, lifeless, janky transition, this looks wrong. NOT for backend logic, browser-test verification (use `qa`), root-cause debugging of a broken render (use `debug`), or throwaway variant exploration before a direction is chosen (use `prototype`).

Bounded fresh-context handover for unfinished mechanical build units from an existing plan. A Claude Code or Codex orchestrator may start a fresh Claude or Codex worker and keeps ownership. - /handover [--agent claude|codex] [slug]: sync handoff through `scripts/verbs fresh-run`. - /handover --async [slug]: write a self-contained payload to docs/handoffs/ only; does not spawn Codex or touch git. Use when a locked, file-scoped task benefits from fresh conversation context. NOT for plan writing, closing finished work, PR/ship flow, or exploratory judgment-heavy work (pull a cross-model take with advisor instead).

Review a code diff before PR or on request. Uses a bounded low-risk fast path, scoped evidence, risk-triggered lenses, and cold-context escalation. NOT browser QA, prepared-plan critique, or external artifact trust checks.

Focused execution from a concrete outcome to SHIPPED, PAUSED, FAILED, or ABORTED_BY_USER. Adds acceptance-driven loops, bounded review, and delivery evidence beyond native coding behavior. Routes unclear requirements to grill, bugs to debug, UI work to ui, and completed delivery to ship. NOT for hypothetical or planning-only responses.

Audit and reduce a live multi-runtime agent harness after adoption. Use for installed skill/plugin parity, cold-start context, routing overlap, telemetry, or human-attention load. Produces a reversible proposal; it does not mutate the harness.

Adversarial requirement discovery. Ask ONE question at a time, hunting for hidden requirements / unknown unknowns. Drills one angle at a time, then by default runs a structured close that writes a brief (+ executable plan) to docs/. Say "quick" or "don't write files" to stop at a chat-only log instead. Use when the user says "grill me on X", "interrogate this idea", "stress test this scope", "what am I missing", "draft a brief", "structured intake". Skip for tasks where scope is already concrete.

Chart or work a decision map across sessions until the route to its destination is clear. With a large, fuzzy topic and no map, delegate charting to grill and stop after the map is created. With an existing map, take ONE unblocked entry, resolve it by type (research / grilling / prototype / task), write the decision back, and graduate the fog. Use when starting or resuming a multi-session effort. NOT executing a locked plan (sprint).

Throwaway prototype that answers ONE design question. Logic branch ("does this state model / data shape feel right?") → tiny interactive terminal driver. UI branch ("what should this look like?") → several structurally different variants behind a ?variant= param. Use when a design question is answerable by building cheap and reacting: sanity-check a state model, explore what a surface should look like, make a grill open question concrete. NOT committed production UI work (use `ui`) and NOT the feature itself (use `sprint`).

Systematic root-cause debugging. Use for errors, crashes, regressions, failing tests, "used to work and now doesn't", or flaky/intermittent behavior. Triggers: 排查, 查 bug, 報錯, 崩潰, 跑不通, 以前是好的, 回歸, 反覆修不好, debug, why is this broken, regression, root cause, used to work, stack trace. NOT for code-diff review (use `review`), UI taste complaints (use `ui`), or building new features.

Browser-based QA. Use when UI has changed, or when asked to "test this", "QA", or "check the page". Requires browser automation from the host. NOT for non-UI checks (use the project's normal test or verification path); NOT for code-diff review (use `review`).

Close completed code work through test, commit, push, and PR. Use when asked to "ship", "create PR", or publish the current branch. To hand unfinished work to Codex, use handover.

Use when working on production code, shared infrastructure, or unfamiliar codebases. Adds confirmation gates before destructive commands (force push, rm -rf, publish, DROP).

Deep-module design vocabulary: a lot of behaviour behind a small interface, placed at a clean seam, testable through that interface. Use when designing or reshaping a module's interface, deciding where a seam goes, judging shallow vs deep, making code testable, or when another skill needs these terms. NOT diff review (use `review`) or requirement discovery (use `grill`).

Pre-adoption trust check for software artifacts before they touch your system: skill/MCP installs, GitHub repositories, URLs/documents, packages, APIs, SDKs, and software services. Triggers on /gatekeeper, /slowmist-agent-security (alias), "is this safe to install", "check this repo".