slate-ar-next

Selects and runs one safe next step for Slate v2 autoresearch workflows.

16.5k|997|Updated Dec 2, 2019
One-click install
npx skills add https://github.com/udecode/plate --skill slate-ar-next
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: slate-ar-next
Source: https://github.com/udecode/plate/tree/main/.agents/skills/slate-ar-next
Command: npx skills add https://github.com/udecode/plate --skill slate-ar-next

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

When iterating on Slate v2 autoresearch work, developers must constantly decide which specialized workflow (status, gate, patch, perf, quality, finalize) to run next. This Skill removes that decision burden by inspecting the current state and automatically choosing and executing one safe next step.

Core Features & Use Cases

  • Automatic owner selection: Reads current status, identifies the strongest blocker or opportunity, and routes to exactly one appropriate owner such as slate-patch, slate-ar-gate, slate-ar-perf, or slate-ar-quality.
  • Single safe step execution: Runs only one meaningful step per invocation, then reports the metric or check result plus the next recommended owner.
  • Escalation routing: Detects when work is too broad for one step and hands off to slate-ar-perfect, slate-ar-fast, or slate-ar-stabilize for durable loops.
  • Use Case: A developer working on Slate v2 editor improvements runs the command each morning; the Skill checks status, finds a known correctness failure, runs one slate-patch step, and reports the result with the next recommended action.

Quick Start

Ask the assistant to run slate-ar-next to inspect the current Slate v2 autoresearch state and perform the single most useful next step.

Frequently Asked Questions about slate-ar-next

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I decide which Slate v2 autoresearch step to run next?

Run slate-ar-next, which reads the current status, identifies the strongest blocker or opportunity, and chooses exactly one owner such as slate-patch, slate-ar-gate, or slate-ar-perf. It then executes one safe step and reports the result.

What is the difference between slate-ar-next and slate-ar-perfect?

slate-ar-next performs a single safe step for daily driving, while slate-ar-perfect handles broad, multi-step surface improvement loops. When the state is too broad for one step, slate-ar-next recommends or hands off to slate-ar-perfect.

Can slate-ar-next make multiple changes in one run?

No, it deliberately stops after the first meaningful packet or fix unless the user explicitly requests a durable loop. This keeps each invocation safe and reviewable.

When does slate-ar-next route to slate-patch versus slate-ar-gate?

Known correctness failures route to slate-patch, while existing behavior proofs that need repeatability route to slate-ar-gate. Missing behavior oracles route to slate-patch or tdd.

What does slate-ar-next report after running a step?

It reports the active session and target, the owner chosen and why, the one step performed, the metric or check result, and the next recommended owner or blocker.