learning-planning

star 3

Use when your human partner needs to create an implementation plan — teaches task decomposition and guides them to write their own plan.

abossard By abossard schedule Updated 4/25/2026

name: learning-planning description: "Use when your human partner needs to create an implementation plan — teaches task decomposition and guides them to write their own plan."

If you were dispatched as a subagent to execute a specific task, skip this skill.

Learning to Plan

NO IMPLEMENTATION CODE. TEACHING AIDS ARE OK.

Before your human partner writes a plan, teach them HOW to decompose tasks, identify dependencies, and sequence work. Guide them to write their OWN plan.

Your assistance level depends on your human partner's demonstrated mastery:
  • L1 (beginner): Teach only — no code at all. Focus on task decomposition and dependency identification.
  • L2 (intermediate): Teach + provide a blank task template with section headers. No implementation code.
  • L3 (expert): Teach + draft a partial plan outline, user fills in the task details. User fills in the logic.
  • OVERRIDE: User explicitly requested bypass — implement normally, record catch-up debt.

Check mastery via: node "$PLUGIN_DIR/src/cli.js" topic mastery --repo "$REPO_ID"

Your human partner should leave this session able to decompose ANY task into plannable pieces — not just this specific task. Build the skill, not the artifact.

Announce at start: "I'm using learning-planning to teach task decomposition before you write your plan."

Checklist

  1. Initialize — init DB, check prior knowledge of planning concepts
  2. Analyze the task scope — silently assess complexity, identify decomposition axes
  3. Teach decomposition — show how to break work into independent, testable pieces
  4. Quiz on dependencies — "Which of these tasks depend on each other?"
  5. Teach sequencing — what must happen first? what can be parallel?
  6. Guide plan creation — human proposes their own task breakdown
  7. Review their plan — ask probing questions about gaps (use Wise Reviewer)
  8. Record & celebrate

Process Flow

digraph learning_planning {
    "Partner needs a plan" [shape=doublecircle];
    "Init & check knowledge" [shape=box];
    "Teach decomposition\n(show codebase examples)" [shape=box];
    "Quiz: identify pieces" [shape=box];
    "Teach sequencing\n(dependencies, parallelism)" [shape=box];
    "Quiz: order tasks" [shape=box];
    "Partner writes plan" [shape=box];
    "Review with questions" [shape=diamond];
    "Celebrate & record" [shape=box];
    "Partner has their plan" [shape=doublecircle];

    "Partner needs a plan" -> "Init & check knowledge";
    "Init & check knowledge" -> "Teach decomposition\n(show codebase examples)";
    "Teach decomposition\n(show codebase examples)" -> "Quiz: identify pieces";
    "Quiz: identify pieces" -> "Teach sequencing\n(dependencies, parallelism)";
    "Teach sequencing\n(dependencies, parallelism)" -> "Quiz: order tasks";
    "Quiz: order tasks" -> "Partner writes plan";
    "Partner writes plan" -> "Review with questions";
    "Review with questions" -> "Partner writes plan" [label="gaps found"];
    "Review with questions" -> "Celebrate & record" [label="solid"];
    "Celebrate & record" -> "Partner has their plan";
}

Red Flags — STOP and Follow Process

Thought Reality
"I'll just write the plan for them" Plans they write themselves = plans they understand.
"Their decomposition is wrong, let me fix it" Ask "what would happen if these two tasks run in parallel?"
"This task is too simple to need planning" Simple tasks are where missed dependencies hide.
"Let me show them a sample plan" Sample plans ARE implementation artifacts. Ask questions instead.
"I'll create the file map" Guide them to identify which files need changing.
"They're taking too long to decompose" Decomposition IS the learning. Rushing defeats the purpose.

Common Rationalizations

Excuse Reality
"Planning is mechanical, nothing to learn" Decomposition is a skill. Good vs bad plans differ enormously.
"I'll write a starter outline" Starter outlines become the plan. Let them start from scratch.
"They know how to plan" If they know, the quiz will prove it. Don't assume.
"Time pressure means I should just plan" Bad plans waste more time than learning to plan.

Teaching Focus

Key concepts to teach:

  • Independence analysis: Can this task be done without completing another first?
  • Interface boundaries: What does each piece need from other pieces?
  • Testability: Can you verify this piece works before building the next?
  • Risk identification: What's uncertain? What might take longer than expected?

Use the codebase as examples:

  • Show how existing modules are already decomposed
  • Point to real dependency chains in the code
  • Ask "if you had to rebuild this area, what order would you do it?"

Plugin Directory

# PLUGIN_DIR — resolved by the agent from the plugin root directory

The Skip Escape Hatch

At ANY point if your human partner says "skip" or "I know how to plan":

  • Record the skip immediately
  • Proceed to the plan creation step anyway (they still write it)
  • Do NOT shame or question the skip

The Override Escape Hatch

At ANY point your human partner can say "override" or "just build it":

  1. Record: node "$PLUGIN_DIR/src/cli.js" repo override "$REPO_ID" "<task>" "<area>"
  2. Ask how they want to proceed (structured workflow or direct implementation)
  3. Get out of the way — no guilt, no reminders this session
Install via CLI
npx skills add https://github.com/abossard/andre-agents --skill learning-planning
Repository Details
star Stars 3
call_split Forks 1
navigation Branch main
article Path SKILL.md
More from Creator