📚 SkillLens: Transform PR Feedback into Developer Growth

Turn every code review into a learning opportunity. SkillLens automatically analyzes PR review feedback and suggests relevant learning resources from Hyperskill — no configuration, no API keys, just actionable education when you need it most.

graph LR
    A[PR Review<br/>Comments] -->|Analyze| B[SkillLens]
    B -->|AI-Powered<br/>Topic Extraction| C[Learning Topics]
    C -->|Match with<br/>Hyperskill| D[Curated<br/>Resources]
    D -->|Post as<br/>Comment| E[📚 Learning<br/>Recommendations]

    style B fill:#7B68EE,stroke:#5a4abf,stroke-width:3px,color:#fff
    style E fill:#90EE90,stroke:#228B22,stroke-width:2px

Loading

✨ Why SkillLens?

• ✅ Zero Configuration — Works out of the box with your existing GitHub workflow
• 🔒 Privacy-First — Only reads review comments, never your source code
• 🎯 AI-Powered Intelligence — Contextual understanding, not keyword matching
• 📚 Curated Learning Paths — Direct links to relevant Hyperskill courses and topics
• 🤖 Non-Intrusive — Single, auto-updating comment per PR (no notification spam)
• ⚡ Lightning Fast — Typical execution under 10 seconds

Perfect for teams building a culture of continuous learning, open-source projects supporting contributor growth, and developers seeking personalized skill development.

🚀 Quick Start

Add this workflow to .github/workflows/skilllens.yml in your repository:

name: SkillLens Learning Recommendations

on:
  pull_request_review:
    types: [submitted, edited]
  issue_comment:
    types: [created, edited]

permissions:
  pull-requests: write
  id-token: write

concurrency:
  group:
    skilllens-${{ github.event.pull_request.number || github.event.issue.number }}
  cancel-in-progress: false

jobs:
  skilllens:
    if:
      ${{ github.event_name != 'issue_comment' || github.event.issue.pull_request }}
    runs-on: ubuntu-latest
    steps:
      - uses: hyperskill/skilllens-action@v1
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

That's it! 🎉 The action will now automatically post learning recommendations when reviewers provide feedback.

📖 How It Works

1. Monitors PR reviews, inline comments, and conversation threads
2. Filters noise (LGTM, emojis) to focus on substantive feedback
3. Analyzes comments using AI to identify learning opportunities
4. Matches topics with Hyperskill's comprehensive course catalog
5. Posts a single, updateable comment with personalized recommendations

See an Example Output

When a reviewer comments:

"Consider using list comprehension instead of a for loop here"

SkillLens posts:

---

📚 Learning Recommendations from Hyperskill

Based on recent review feedback, here are personalized learning resources:

1. Python Collections & List Comprehensions (Confidence: 92%) Learn efficient list operations and Pythonic patterns 📖 View Course → · 🎯 Practice Problems →

2. Code Optimization Techniques (Confidence: 87%) Master performance optimization in Python applications 📖 View Course → · 📚 Learning Track →

---

_{🤖 Generated from review comments only • No source code analyzed • Learn more}

⚙️ Configuration

Inputs

Input	Description	Required	Default	Example
oidc-audience	OIDC audience for backend authentication	No	skilllens.dev	custom.domain.com
default-language	Fallback language when repository language unknown	No	Python	JavaScript, Java
max-topics	Maximum learning topics to include	No	5	3, 10
min-confidence	Minimum confidence threshold (0-1)	No	0.65	0.5, 0.8
fail-on-proxy-error	Fail workflow on backend errors	No	false	true

Outputs

Output	Description	Example Use
topics-json	JSON array of identified learning topics	Store for metrics/analysis
comment-url	URL of the created/updated PR comment	Send notifications

📝 Advanced Configuration Examples

Language-Specific Configuration

# Python Projects
- uses: hyperskill/skilllens-action@v1
  env:
    GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
  with:
    default-language: Python
    max-topics: 5

# JavaScript/TypeScript Projects
- uses: hyperskill/skilllens-action@v1
  env:
    GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
  with:
    default-language: JavaScript
    min-confidence: 0.7

# Java Projects
- uses: hyperskill/skilllens-action@v1
  env:
    GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
  with:
    default-language: Java
    fail-on-proxy-error: true

Using Outputs

- name: Generate Recommendations
  id: skilllens
  uses: hyperskill/skilllens-action@v1
  env:
    GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

- name: Save Topics
  run: echo "${{ steps.skilllens.outputs.topics-json }}" > topics.json

- name: Notify Slack
  if: steps.skilllens.outputs.comment-url
  run: |
    curl -X POST ${{ secrets.SLACK_WEBHOOK }} \
      -d "text=New learning recommendations: \
      ${{ steps.skilllens.outputs.comment-url }}"

🔒 Privacy & Security

Your code stays yours. SkillLens is built on privacy-first principles.

What We Access

✅ PR review comments and feedback ✅ Repository metadata (name, PR number) ✅ Public PR conversation threads

What We DON'T Access

❌ Your source code ❌ File contents ❌ Private repository data ❌ Secrets or credentials

Authentication

• Uses GitHub's secure OIDC tokens (no API keys required)
• Tokens are short-lived and scoped to the workflow
• Backend verification ensures only legitimate GitHub Actions can call our API

Required Permissions

Permission	Level	Purpose
pull-requests	write	Read PR review comments and create/update comments
id-token	write	Request OIDC authentication token

Note: The pull-requests: write permission typically covers both reading PR data and writing comments. If you encounter permission issues, see the troubleshooting section for alternative configurations.

🎯 Use Cases

For Development Teams

Build a culture of continuous learning by automatically connecting code review feedback to educational resources. Perfect for onboarding, skill development, and knowledge sharing.

For Open Source Projects

Support contributor growth by providing learning resources alongside review feedback. Reduce back-and-forth by helping contributors understand the "why" behind suggestions.

For Individual Developers

Turn every PR into a personalized learning opportunity. Never wonder where to learn a concept mentioned in reviews — get instant, relevant recommendations.

For Engineering Managers

Track skill development trends across your team. Identify common learning needs and inform training priorities based on actual code review patterns.

🤝 Comparison with Alternatives

Feature	Manual Learning Search	AI Code Reviewers	SkillLens
Automated	❌ Manual effort	✅ Yes	✅ Yes
Learning-Focused	⚠️ If you remember	❌ Code quality only	✅ Primary focus
Privacy	✅ No tool needed	❌ Reads source code	✅ Reviews only
Setup Complexity	N/A	🔧 API keys required	✅ Zero config
Cost	Time-consuming	💰 Usually paid	🆓 Free
Noise Level	N/A	📢 Multiple comments	🔇 Single comment

❓ Frequently Asked Questions

Does SkillLens read my source code?

No. SkillLens only analyzes PR review comments and feedback. It never accesses your repository's source code, ensuring complete privacy and security.

What programming languages are supported?

SkillLens supports all languages available on Hyperskill, including:

• Python, Java, JavaScript, TypeScript
• Kotlin, Go, Scala, C++
• SQL, Bash, and more

The default-language input helps optimize recommendations when the repository language cannot be automatically detected.

Can I use this with private repositories?

Yes! SkillLens works with both public and private repositories. It uses GitHub's built-in authentication and only accesses review comments, which are already visible to workflow runs.

How accurate are the recommendations?

SkillLens uses advanced AI to understand context and nuance in review comments. The default confidence threshold of 0.65 ensures only relevant recommendations are shown. You can adjust this with the min-confidence input.

Will this spam my PRs with comments?

No. SkillLens maintains a single comment per PR that updates in place. It uses a hidden HTML marker to identify and update the same comment, preventing notification spam.

What if the SkillLens API is down?

By default, the action logs a warning but doesn't fail your workflow. You can change this behavior by setting fail-on-proxy-error: true.

🚨 Troubleshooting

Common Issues and Solutions

"Resource not accessible by integration"

Error message:

Error: Resource not accessible by integration

Cause: Insufficient GitHub token permissions for accessing PR data or creating comments. This typically occurs when the workflow lacks proper permissions for the repository.

Solution 1: Ensure you have the minimal required permissions:

permissions:
  pull-requests: write
  id-token: write

Solution 2: If the minimal permissions don't work due to repository settings, try this expanded configuration:

permissions:
  contents: read # Repository metadata access
  pull-requests: write # PR review and comment access
  issues: read # Additional PR comment read access
  id-token: write # OIDC authentication

Note: Repository settings can vary, and some organizations may require additional permissions depending on their security policies. For organization-owned repositories, check that "Allow GitHub Actions to create and approve pull requests" is enabled in Settings → Actions → General → Workflow permissions.

"Unable to get ACTIONS_ID_TOKEN_REQUEST_URL env variable"

Error message:

Error message: Unable to get ACTIONS_ID_TOKEN_REQUEST_URL env variable

Cause: Missing id-token: write permission. GitHub Actions only provides the OIDC token environment variables when this permission is explicitly granted.

Solution: Add the id-token: write permission to your workflow:

permissions:
  pull-requests: write
  id-token: write # Required for OIDC authentication

Without id-token: write, the action cannot authenticate with the SkillLens backend and will fail immediately after fetching review data.

"No comment appears after reviews"

Possible causes:

1. No actionable feedback (filtered noise like "LGTM")
2. Confidence threshold not met
3. API communication issue

Debug steps:

1. Enable debug logging using one of these methods:
   • Re-run the job with debug logging from GitHub UI (click "Re-run jobs" → "Enable debug logging")
   • Add a secret named ACTIONS_STEP_DEBUG with value true to your repository
2. Lower the confidence threshold for testing:

- uses: hyperskill/skilllens-action@v1
  with:
    min-confidence: 0.5 # Lower threshold for testing

"Action runs but fails immediately"

Check:

1. Workflow permissions are correctly set
2. Action is triggered on correct events
3. Repository allows GitHub Actions to create comments

For organizations: Go to Repository Settings → Actions → General → Workflow permissions → Enable "Allow GitHub Actions to create and approve pull requests"

🛠️ Development

Local Development Setup

# Clone the repository
git clone https://github.com/hyperskill/skilllens-action.git
cd skilllens-action

# Install dependencies
npm install

# Run tests
npm run test

# Run all checks (format, lint, test, bundle)
npm run all

# Test locally with stubbed GitHub Actions environment
npm run local-action

Important: Bundle Before Commit

After making changes to TypeScript sources in src/:

npm run bundle  # Regenerates dist/index.js

The dist/ directory must be committed with your changes.

📊 Roadmap

• Support for custom learning platforms beyond Hyperskill
• Team-level skill gap aggregation dashboard
• IDE integrations (Visual Studio Code, JetBrains)
• Multi-language UI support
• Webhook support for external integrations
• Historical learning trends analysis

👥 Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

Quick contribution checklist:

• Fork and create a feature branch
• Make changes to TypeScript files in src/
• Run npm run all to ensure all checks pass
• Update dist/ with npm run bundle
• Submit PR with clear description

🙏 Support

• 📖 Documentation
• 🐛 Report Issues
• 💡 Request Features
• 📧 Email: skilllens@hyperskill.org

📄 License

MIT License - see LICENSE for details.

🌟 Acknowledgments

• Built with actions/typescript-action
• Powered by Hyperskill
• Uses GitHub OIDC for secure authentication

---

⭐ Star this repository if SkillLens helps your team learn and grow!

📢 Share your experience with the community using #SkillLens