make-plan

Creates phased implementation plans grounded in documentation discovery and subagent research.

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/LGJ-Jonathan/Jonathan_Global_Claude_Skills --skill make-plan-lgj-jonathan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: make-plan
Source: https://github.com/LGJ-Jonathan/Jonathan_Global_Claude_Skills/tree/main/claude-mem/plugin/skills/make-plan
Command: npx skills add https://github.com/LGJ-Jonathan/Jonathan_Global_Claude_Skills --skill make-plan-lgj-jonathan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When planning a feature or multi-step implementation, AI agents often invent APIs, assume undocumented parameters, or skip verification, producing plans that fail during execution. This Skill enforces a documentation-first planning workflow so every phase cites real sources and includes verification steps. ## Core Features & Use Cases - Documentation Discovery Phase: Deploys subagents to find actual APIs, signatures, and examples before any implementation planning begins. - Subagent Reporting Contract: Requires every research subagent to cite sources, concrete findings, copy-ready snippet locations, and confidence notes. - Phased Plan Structure: Produces self-contained phases with documentation references, verification checklists, and anti-pattern guards, executable in fresh chat contexts. - Use Case: Before building a new API integration, use this Skill to generate a plan where each phase points to exact documentation lines to copy from, with grep checks and tests proving each phase worked. ## Quick Start Ask the AI to create a phased implementation plan for your feature using the make-plan skill before writing any code.

Frequently Asked Questions about make-plan

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

FAQPage Schema
How do I create an implementation plan before coding a feature?

Start with a documentation discovery phase that identifies the actual APIs and signatures available, then break work into self-contained phases. Each phase should cite specific documentation lines to copy from and include a verification checklist.

How to prevent AI agents from inventing APIs that don't exist?

Require a documentation discovery step where subagents read real docs and produce an Allowed APIs list with cited sources. Reject any subagent report that states conclusions without listing the files or URLs consulted.

What should each phase of an implementation plan include?

Each phase needs a task framed around copying from documentation, specific file and line references, a verification checklist with tests or grep checks, and anti-pattern guards listing what not to do.

When should I use a planning skill instead of coding directly?

Use it for multi-step features or tasks executed across separate chat sessions, where each phase must be self-contained. For trivial single-file changes, direct implementation is usually faster than a phased plan.

Why do AI-generated implementation plans fail during execution?

Plans fail when they assume APIs exist without checking documentation, add undocumented parameters, or skip verification steps. Grounding every phase in cited sources and requiring proof of completion prevents these failures.