planning

Creates durable, resumable multi-step plans as Markdown files with goals, work units, and progress trackers.

Updated Aug 4, 2026
One-click install
npx skills add https://github.com/tschallacka/ai-skills --skill planning-tschallacka
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: planning
Source: https://github.com/tschallacka/ai-skills/tree/main/planning
Command: npx skills add https://github.com/tschallacka/ai-skills --skill planning-tschallacka

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Multi-step initiatives lose context when a conversation ends or an agent session resets. This Skill turns an initiative into a directory of Markdown plan files — goals, ordered steps, work units, verification instructions, progress trackers, and handoff notes — that another agent can resume and execute without reconstructing missing context. ## Core Features & Use Cases - Plan decomposition: Creates a plan directory with a plan description, work-unit inventory, goals, and step files, each written through validated helper scripts. - Adversarial review gates: Requires an independent reviewer pass with findings, fix-key verification, and an approval gate before a plan counts as ready. - Progress tracking and resumption: Generates plan-level and goal-level progress trackers plus bounded context snapshots so work can resume across sessions. - Use Case: Ask the agent to plan a multi-week refactor; it produces a validated plan directory with ordered goals and steps, passes adversarial review, and later sessions resume execution from the trackers. ## Quick Start Use the planning skill to create a durable, resumable plan for migrating our billing service to the new API, with goals, ordered steps, and progress trackers.

Frequently Asked Questions about planning

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

FAQPage Schema
How do I create a resumable multi-step plan for a coding agent?▼

Run the planning skill's create-plan.sh helper to generate a plan directory with a plan description, work-unit inventory, and command registry, then add goals and work units through the provided helpers. The result is a set of Markdown files another agent can resume without rebuilding context.

How does the planning skill verify a plan is ready to execute?▼

A plan passes through an independent adversarial review, fix-key verification, and the validate-plan.sh readiness gate before progress trackers are created. The strict --complete form promotes warnings to failures, so only fully validated plans ship.

When should I not use a durable plan file?▼

Do not use this skill for small, self-contained changes or temporary in-chat checklists. It is designed for multi-step initiatives that genuinely need resumable files, ordered goals, verification instructions, and handoff notes.

Can a plan be resumed by a different agent or session?▼

Yes. The skill writes bounded context snapshots, progress trackers, and handoff notes under the plan directory so a fresh agent can read scoped slices and continue execution. Role-gated readers cap how much plan content each agent loads.

Why does plan validation fail on placeholders or unregistered commands?▼

The validator fails when template placeholders remain in generated artifacts or when command literals in step files are not registered in the plan's commands.json. Register commands with register-command.sh and replace placeholders through the update helpers.