Curriculum Designer

Design customized curricula for Apni Pathshala PODs with real YouTube video links.

FEATURES:

• - ✅ Staged execution with checkpointing (recovery from failures)
• ✅ YouTube link verification with fallback logic (no blank URLs)
• ✅ Context capping per lesson (reduced token usage)
• ✅ Every topic gets a valid video OR search query fallback

---

⚡ Quick Start

How This Skill Works

When invoked, the agent follows a 5-stage workflow with checkpointing:

Stage	What Happens	Checkpoint File
1	Gather requirements	INLINECODE0
2

Research YouTube videos | research-results.json |
| 3 | Verify videos + fallback logic | validated-resources.json |
| 4 | Design curriculum (one lesson at a time) | curriculum-structure.json |
| 5 | Create Google Sheet | final-sheet-url.txt |

Checkpoint Behavior

• - Each stage saves its output to a checkpoint file
• If checkpoint exists, stage loads it and skips processing
• If checkpoint doesn't exist, stage runs from scratch
• Re-running resumes from first incomplete stage

---

Trigger

User message contains:

• - "Design curriculum" → Start curriculum creation
• "Create curriculum for [POD name]" → Start with POD context
• "Build learning plan" → Start curriculum creation
• "Curriculum for [subject/topic]" → Start with topic context

---

Target User

This skill is designed for Madhur (Academic Associate) who designs curricula for PODs.

---

Configuration

• - API Keys: Stored locally in ~/.openclaw/workspace/skills/curriculum-designer/.env (NOT in git)
• Output Folder: 1upJQu-IVmZRJQsNGmJNRzq9IwL67MVL9 (Curriculum Designer)
• Checkpoint Directory: INLINECODE7

YouTube API Key:
CODEBLOCK0

Get from: https://console.cloud.google.com/apis/credentials

---

Agent Workflow Instructions

Model Allocation for Stages

Action: Use different LLM models for different stages to optimize cost and performance.

Stage-Specific Model Assignment

Stage	Recommended Model	Reason
Stage 1: Requirements Collection	glm-4.7	Quick reasoning, sufficient for form filling
Stage 2: YouTube Research

glm-5 | Fast research, needs latest web knowledge | | Stage 3: Video Validation | glm-4.7 | Pattern matching, simple logic | | Stage 4: Curriculum Design | glm-4.7 | Structured generation, cost-effective for lessons | | Stage 5: Sheet Creation | glm-4.7 | JSON formatting, simple transformations |

How to Set Models

Option 1: Specify model when calling agent
CODEBLOCK1

Option 2: Configure in SKILL.md
Each stage should include model recommendation in its instructions:

CODEBLOCK2

---

Agent Workflow Instructions

Stage 1: Gather Requirements

Action: Ask the user these questions (from SOP):

Basic Information

1. 1. POD Name - Which POD is this curriculum for?
2. Target Audience - Grade level or age group of students?
3. Subject Areas - What subjects/topics should be covered?
4. Duration - How long is the program? (e.g., 1 month, 3 months, 6 months)
5. Frequency - How many classes per week?
6. Daily Lab Hours - How many hours will the lab operate?
7. Previous Exposure - Have students done digital learning before?

Teacher Context

1. 8. Teacher Capability - Can teachers operate computers independently?
2. Teacher Training Needed - Do teachers need any training?

Learning Outcomes

1. 10. Learning Area Focus - Which area(s) to prioritize?

- Digital Literacy - Academic Empowerment - Skill Development - Employment Readiness

1. 11. Specific Skills - What specific skills should students acquire?
2. Assessment Method - How will learning be measured?

Output: Save to checkpoint as JSON:
CODEBLOCK3

Checkpoint: ~/.openclaw/workspace/curriculum-designer-checkpoints/<timestamp>-<session-id>/requirements.json

---

Stage 2: Research YouTube Resources

Action: Search YouTube for videos based on requirements

API: Use YouTube Data API v3 with key from INLINECODE9

Search Queries (Default):
CODEBLOCK4

Search Parameters:

• - INLINECODE10
• INLINECODE11
• INLINECODE12
• INLINECODE13
• INLINECODE14 (5-10 minutes preferred)
• INLINECODE15 (Hindi preference)

Output Structure:
CODEBLOCK5

Research Summary (Before Validation)

After completing all searches, summarize the research results before passing to validation stage.

Why summarize?

• - Reduces token usage when passing to Stage 3 (validation)
• Provides cleaner input for validation logic
• Allows easy review of what was researched

Summary Structure:
CODEBLOCK6

Save summary:

• - Append research_summary to INLINECODE17
• Validation stage uses summary for context, not raw results

Checkpoint: ~/.openclaw/workspace/curriculum-designer-checkpoints/<timestamp>-<session-id>/research-results.json

---

Stage 3: Verify Videos + Fallback Logic

Action: Verify each video via YouTube oEmbed API. If invalid, retry with alternative search terms.

Verification Method

Use oEmbed endpoint (fast, lightweight):
CODEBLOCK7

• - HTTP 200 = Valid video
• HTTP 404 = Invalid/deleted video
• HTTP 4xx/5xx = Try again (rate limit or temporary error)

Fallback Logic (CRITICAL)

For each topic, follow this logic:

CODEBLOCK8

Output Structure (With Fallbacks)

CODEBLOCK9

IMPORTANT: Every topic MUST have either:

• - A valid video URL, OR
• A search query fallback

Checkpoint: ~/.openclaw/workspace/curriculum-designer-checkpoints/<timestamp>-<session-id>/validated-resources.json

---

Stage 4: Design Curriculum (Context Capping + Summarization)

Action: Generate curriculum structure, processing one lesson at a time with summarization and context cleanup.

How Context Capping + Summarization Works

Instead of:
CODEBLOCK10

Do this:
CODEBLOCK11

Lesson-by-Lesson Process

For lesson N:

1. 1. Load context:

- Lesson N topic - Lesson N resources (from validated-resources.json) - Previous lesson summary (if N > 1)

1. 2. Generate lesson:

- Daily learning objectives - Daily assessment - Module content - YouTube link/fallback

1. 3. Summarize lesson:

- Create concise summary of lesson N - Focus on: key skills, tools used, assessment type

1. 4. Save to curriculum:

- Full lesson details - Lesson summary (for next lesson's context)

1. 5. Context cleanup:

- Remove lesson N's full context from memory - Keep only lesson N's summary for N+1

Lesson Summary Template

CODEBLOCK12

Lesson Generation Template

For each lesson, generate:

Field	Description
Day	Lesson number (1, 2, 3, ...)
Subject

Subject area / Learning area |
| Module | Module/Topic name |
| Daily Learning Objectives | What students learn that day |
| Daily Assessment | How to assess understanding |
| YouTube Link | Valid video URL OR search query fallback |
| YouTube Title | Video title (if applicable) |
| Tools Used | Required software/platforms |
| Fallback Search Query | Search query if no valid video (or blank) |
| Lesson Summary | Concise summary for next lesson's context |

Sample Lesson Output (With Summary)

CODEBLOCK13

With Fallback Example

CODEBLOCK14

Final Curriculum Summary (When All Lessons Complete)

After generating all lessons:

1. 1. Summarize entire curriculum:

- Total lessons - Subject areas covered - Key skills progression - Assessment approach - Tools/software used

1. 2. Save to curriculum structure:

CODEBLOCK15

1. 3. Update Stage 2 checkpoint:

- Add curriculum_summary field to research-results.json - This keeps summary alongside research results for reference

Checkpoint: INLINECODE22

Also updates: ~/.openclaw/workspace/curriculum-designer-checkpoints/<timestamp>-<session-id>/research-results.json (adds curriculum_summary)

---

Stage 5: Create Google Sheet

Action: Create Google Sheet with curriculum data using gog CLI.

Step 1: Create Sheet

CODEBLOCK16

Step 2: Add Headers

CODEBLOCK17

Step 3: Populate with Lessons

CODEBLOCK18

Data format:

• - Column A: Day (1, 2, 3, ...)
• Column B: Subject
• Column C: Module
• Column D: Daily Learning Objectives
• Column E: Daily Assessment
• Column F: YouTube Link
• Column G: YouTube Title
• Column H: Tools Used
• Column I: Fallback Search Query

Validation:

• - One row per lesson
• Include fallback search queries (if any)
• Ensure no blank YouTube Links OR populated Fallback Search Query

Step 4: Share Sheet

CODEBLOCK19

Step 5: Save URL

CODEBLOCK20

Checkpoint: ~/.openclaw/workspace/curriculum-designer-checkpoints/<timestamp>-<session-id>/final-sheet-url.txt

---

Learning Areas Framework

Learning Area	Focus
Digital Literacy	Basic computer skills, internet safety, AI tools
Academic Empowerment

Study skills, exam prep, note-taking |
| Skill Development | Programming, design, content creation |
| Employment Readiness | Resume, communication, job skills |

---

Video Selection Criteria

• - ✅ 5-10 minutes max - keeps engagement high
• ✅ Clear explanations - no jargon-heavy content
• ✅ Hindi or bilingual - accessible for all students
• ✅ Recent content - prefer 2023+ videos
• ❌ Long lectures - students lose interest
• ❌ Advanced content - match to target audience level

---

Important Guidelines

⚠️ CRITICAL: Sharing Permissions

• - ALWAYS share sheet with --to anyone --role reader before returning link
• NEVER return a restricted link - user cannot view it
• Command: INLINECODE26

⚠️ CRITICAL: No Blank URLs

• - Every topic MUST have either:

- A valid YouTube URL, OR - A search query fallback

• - Never leave both fields blank

Assessment Design

• - Formative (daily): Quick quizzes, practice exercises, short tasks
• Summative (end): Projects, presentations, comprehensive tests
• Keep assessments practical and hands-on

---

Folder Reference

Resource	Link
Curriculum Designer Folder	INLINECODE27
Example Curriculum (AI Tools)

https://docs.google.com/spreadsheets/d/1hYC2Q2KlW8dM71biC97RPSvFnxTQa-zN | | SOP Document | https://docs.google.com/document/d/1Y5qetW8S4RWsTg7hycIyujgTwTCFn9VV |

---

Security Note

⚠️ API keys are stored locally in .env file - NEVER commit this file to git!

---

Future Improvements

1. 1. Resume from specific stage - Ability to jump to any stage, not just first failed one

---

Automatic Checkpoint Cleanup (Cron Job)

Purpose

Delete checkpoint directories older than 7 days to prevent disk space bloat while keeping recent sessions for debugging.

Cron Job Configuration

Option 1: Add to User Crontab

CODEBLOCK21

Option 2: Using OpenClaw Cron

CODEBLOCK22

Cron Schedule Options

Schedule	Crontab Format	Description
Daily at midnight	INLINECODE31	Every day at 00:00
Weekly on Sunday

0 0 * * 0 | Every Sunday at 00:00 | | Every 6 hours | 0 */6 * * * | Every 6 hours (may be too frequent) | | Twice daily | 0 0,12 * * * | At 00:00 and 12:00 |

Verification

After setting up cron, verify it's working:

CODEBLOCK23

Manual Cleanup Test

Test the cleanup command manually before setting up cron:

CODEBLOCK24

Notes

• - -mtime +7: Files/directories modified more than 7 days ago
• -type d: Only directories (sessions), not individual files
• -exec rm -rf {} \;: Remove directory and all contents
• Checkpoints are preserved after completion for review, then auto-cleaned after 7 days
• Adjust +7 to a different value if you want different retention period (+3, +14, +30)

---

Notes

• - Always search for REAL video URLs before creating curriculum
• Save curriculum sheets in the designated folder
• Share viewable link at the end
• Consider teacher training needs if curriculum requires new tools
• Checkpoints are preserved after completion for review
• Every topic in final curriculum must have valid video OR search query fallback