new-plan

Writes the Plan tab in plan.html and backfills acceptance criteria traceability for a task.

Updated Apr 17, 2026
One-click install
npx skills add https://github.com/mistakenot/skills --skill new-plan-mistakenot
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: new-plan
Source: https://github.com/mistakenot/skills/tree/main/plugins/planning-workflow/skills/new-plan
Command: npx skills add https://github.com/mistakenot/skills --skill new-plan-mistakenot

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Turning an approved solution into an executable implementation plan is error-prone: phases get sliced by layer instead of vertically, acceptance criteria lose their link to the phases and tests that satisfy them, and historical context from past work is missed. This Skill automates that final planning stage so the plan is ready for phased execution. ## Core Features & Use Cases - Context Enrichment: Spawns a subagent to mine git history and completed task folders for related decisions, then merges findings into context.md and flags drifted file paths. - Vertical-Slice Phase Design: Breaks the solution into atomic execution phases with a walking-skeleton first phase, an explicit dependency DAG, per-step verify checks, and commit-plus-verification endings. - AC Traceability Backfill: Updates each pd-ac card in the Verification tab with the phase numbers and test file paths that cover it. - Use Case: After a solution is approved in plan.html, run this Skill to produce the Plan tab with a mermaid execution sequence, then hand off to /commit-task and /execute-task for implementation. ## Quick Start Write the plan for the current task now that the solution is approved.

Frequently Asked Questions about new-plan

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

FAQPage Schema
How do I turn an approved solution into an implementation plan?▼

Run this Skill after the Solution tab is approved. It reads plan.html and context.md, designs atomic execution phases as vertical slices with a walking-skeleton first phase, and writes the Plan tab with a mermaid dependency diagram.

How to link acceptance criteria to implementation phases and tests?▼

The Skill backfills each pd-ac card in the Verification tab with a phases attribute listing covering phase numbers and a tests attribute listing test file paths from the solution, creating traceability from criteria to implementation and verification.

When should I not use this planning skill?▼

Do not use it when no solution exists yet, since it depends on the Requirements, Verification, and Solution tabs being present in plan.html. Complete the earlier planning stages (new-task, new-solution) first.

Why slice phases vertically instead of by layer?▼

Vertical slicing makes Phase 1 a thin end-to-end path through every layer with one passing check, so integration risk surfaces early. Layer-by-layer slicing defers integration to the end and yields no verification signal until then.

What happens after the Plan tab is written?▼

The Skill hard-stops and presents the Plan tab for review. Once approved, you run /commit-task to commit the task docs and then /execute-task to start implementation in an isolated worktree.