editor-harvest-plan

Converts editor test harvest reports into lane-specific execution plans for Slate v2 or Plate.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

After harvesting test candidates from an upstream editor repository, teams need a structured way to route every harvested row to the right owner lane and turn it into an actionable, reviewable plan without accidentally executing implementation work.

Core Features & Use Cases

  • Lane-Based Row Routing: Classifies every harvest report row as in-lane, out-of-lane, split, duplicate, skip, or unresolved for the slate-v2 or plate owner lanes.
  • Execution-Grade Plan Generation: Produces a docs/plans/*-harvest-plan.md document with coverage mapping, action values, verification commands, confidence scoring, and a pass-state ledger.
  • License-Safe Planning: Enforces fresh-invariant-only wording for behavior-only sources so no upstream test code is copied into versioned output.
  • Use Case: After running a test harvester against TinyMCE, invoke this Skill with the slate-v2 lane to produce a complete plan covering every raw Slate candidate, then hand the accepted plan to the downstream lane skill for execution.

Quick Start

Ask the agent to process the harvest report at docs/editor-test-harvester/tinymce/report.md into a slate-v2 lane execution plan.

Frequently Asked Questions about editor-harvest-plan

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

FAQPage Schema
How do I turn an editor test harvest report into an execution plan?

Invoke the skill with a lane and a report path or repo key, such as a lane argument plus the harvest report location. It reads the report, inventory, and test-index, then writes a lane-specific plan under docs/plans with row accounting and an execution queue.

What is the difference between the slate-v2 and plate lanes?

The slate-v2 lane owns raw Slate substrate behavior like selection, IME, clipboard, and normalization in the .tmp/slate-v2 checkout. The plate lane owns Plate packages, kits, docs, examples, and product behavior in the main repository.

Does this skill execute the tests or modify implementation code?

No, it is strictly a planning and routing skill. It never patches implementation code, tests, examples, or build config; execution is handed to the downstream lane skill only after the user accepts the finished plan.

Can I use it without a completed harvest report?

No, a harvest report with inventory evidence is required first. If the report is missing, you must run or request the editor-test-harvester skill before this planning step can proceed.

How does it handle behavior-only licensed sources?

For behavior-only sources, all versioned plan rows must use fresh local invariant wording with source path provenance only. Copying upstream test code, fixtures, snapshots, or expressive prose into versioned output is prohibited.