Lesson Planner Skill - Quick Guide

This skill for OpenClaw is a simplified version of https://github.com/saniales/ai-lesson-planner by Alessandro Sanino that offers extended features like course competitive analysis, but requires GitHub Actions and therefore Copilot. This version allows the use of different models and OpenRouter keys.

Installation

1. In the skill folder:

cd /path/to/lesson-planner-skill
npm install

2. Configure API key:
   • OpenAI: export OPENAI_API_KEY="sk-..."
   • OpenRouter (recommended for more model choices): export OPENROUTER_API_KEY="sk-or-..." Optional: OPENROUTER_REFERER (default: https://openclaw.ai) and OPENROUTER_TITLE (default: OpenClaw Lesson Planner). You can also use a .env file with these variables.

Override endpoint

Use --base-url <url> to specify a custom API endpoint (e.g., for providers compatible with OpenAI). When using OpenRouter, the endpoint is automatic.

3. Install the binary globally (optional):

npm link

Now lesson-planner will be available everywhere.

Quick usage

# 1) Plan a course (in the desired directory)
lesson-planner plan-course \
  --title "Introduction to Python" \
  --audience "high school seniors" \
  --lessons 12 \
  --topics "basic syntax, data types, functions, OOP, files" \
  --lang it \
  --output-dir ./lessons

# 2) Scaffold all lessons (in the same output directory)
lesson-planner scaffold-course --output-dir ./lessons

# 3) Generate content for each lesson (one at a time)
lesson-planner generate-lesson \
  --number 01 \
  --id "intro-python" \
  --title "Introduction to Python" \
  --lang it \
  --output-dir ./lessons

# 4) (optional) Export PDF (all lessons)
lesson-planner export-pdf --output-dir ./lessons

# or single lesson
lesson-planner export-pdf --lesson "intro-python" --output-dir ./lessons

Notes

• Use --output-dir to specify the output directory for course files.
• Alternatively, you can set the LESSON_PLANNER_CONFIG_DIR environment variable.
• Default model: gpt-4o-mini (OpenAI). With OpenRouter use namespaced models like openai/gpt-4o-mini or anthropic/claude-3.5-sonnet.
• This skill is independent of OpenClaw; you can also use it from the command line.

Integration with OpenClaw

Once installed, the skill will be available as an OpenClaw command if registered in the skill manifest. Consult the OpenClaw documentation on how to add custom skills.