Inspiration

Many learners use Slack for communication, but they still need to search for resources, wait for mentor support, and decide what to learn next.

We created SkillBridge Agent to provide structured, AI-powered learning support directly inside Slack.

What it does

SkillBridge Agent acts as an AI learning mentor inside Slack. It helps learners:

  • Check their current skill level
  • Create personalized learning paths
  • Ask technical questions
  • Receive assignments and practice activities
  • Prepare for interviews
  • Continue learning through Slack threads
  • Subscribe to quizzes and learning reminders

Users can interact with the agent through Slack mentions, direct messages, slash commands, modals, and assistant prompts.

For example:

@SkillBridge Agent I want to learn Apex or: /skillbridge I want to become a Salesforce developer

How we built it

We built SkillBridge Agent using Node.js, JavaScript, and the Slack Bolt framework.

The application uses:

  • Slack Events API
  • Slack Socket Mode
  • Slack App Home
  • Slash commands
  • Block Kit modals
  • Interactive feedback buttons
  • Slack assistant threads
  • Gemini API for AI-generated learning support
  • JSON-based storage for learning progress and quiz subscriptions

The app listens for Slack events such as app mentions, direct messages, new channel members, App Home activity, and assistant thread creation.

We also added conversation memory so learners can continue their learning journey within the same Slack thread.

Challenges we ran into

One major challenge was handling the different types of Slack interactions. App mentions, slash commands, direct messages, modals, and feedback buttons all use different event structures.

We also needed to:

  • Acknowledge Slack interactions quickly
  • Prevent the bot from responding to every channel message
  • Maintain separate conversation histories for different users and threads
  • Configure the correct Slack scopes and permissions
  • Handle Gemini API errors and usage limits
  • Keep the agent focused only on learning and upskilling topics

Accomplishments that we're proud of

We are proud that SkillBridge Agent provides a complete learning experience without requiring users to leave Slack.

The project supports:

  • Personalized learning paths
  • Skill assessments
  • Mentor session planning
  • Thread-based conversation memory
  • Direct message support
  • Channel-specific welcome messages
  • Interactive feedback
  • Quiz subscriptions
  • Scheduled learning nudges
  • Learning-focused guardrails

We are also proud that the same application can support both learners and mentors.

What we learned

This project helped us understand how to build an event-driven Slack application and connect multiple Slack features into one useful product.

We learned how to:

  • Build Slack apps with Bolt for JavaScript
  • Work with Slack events and permissions
  • Create modals using Block Kit
  • Implement slash commands and interactive buttons
  • Maintain conversation context
  • Integrate a generative AI API
  • Design learning-focused AI guardrails
  • Test Slack workflows from end to end

We also learned that a useful AI mentor must do more than answer questions. It should provide structure, practical exercises, clear next steps, and continuous learning support.

What's next for SkillBridge Agent

Our next goal is to make SkillBridge more personalized and measurable.

Planned improvements include:

  • Persistent learner profiles
  • Skill-progress dashboards
  • Learning streaks and badges
  • Mentor analytics
  • Trailhead integration
  • Assignment submission and review
  • Smarter quiz generation
  • Cohort-level learning reports
  • Workspace learning insights
  • More advanced Slack MCP integration

Built With

Share this project:

Updates