Your Hexabot instance now has:
- β GitHub MCP Server - Connects AI to GitHub APIs
- β GitHub Assistant Bot Workflow - Conversational AI with GitHub tools
- β
Slack Channel Package -
[email protected]installed - π Setup Guides - Complete documentation for Slack integration
| File | Purpose |
|---|---|
WORKFLOW_SUMMARY.md |
Details of the GitHub Assistant Bot workflow |
SLACK_SETUP_GUIDE.md |
Complete step-by-step Slack integration guide |
SLACK_QUICK_START.md |
5-minute quick start guide for Slack |
setup-slack.sh |
Interactive setup script (helper tool) |
README_INTEGRATION.md |
This file - overview of everything |
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β Hexabot System β
β β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ β
β β Workflows β β
β β ββ GitHub Assistant Bot (published) β β
β β β ββ AI Agent with GitHub MCP binding β β
β β ββ Hello Workflow (for testing) β β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ β
β β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ β
β β MCP Servers β β
β β ββ GitHub MCP (enabled) β β
β β ββ Command: npx @modelcontextprotocol/... β β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ β
β β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ β
β β Channels (Ready to Configure) β β
β β ββ Slack (hexabot-channel-slack installed) β β
β β ββ Waiting for: credentials + source setup β β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ β
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
The workflow is ready to use, but needs a channel to receive messages:
# Via Hexabot Console (if available)
# Or via API/UI when Slack is connectedFollow one of these guides:
- Quick:
SLACK_QUICK_START.md(5 minutes) - Detailed:
SLACK_SETUP_GUIDE.md(complete walkthrough) - Interactive: Run
./setup-slack.sh
Once Slack is connected, users can:
Ask about GitHub:
User: @Hexabot what are the latest releases of hexastack/hexabot?
Bot: [Uses GitHub MCP] Here are the latest releases...
Get Repository Info:
User: Show me the README for anthropics/anthropic-sdk-python
Bot: [Fetches from GitHub] Here's the repository information...
Search Issues/PRs:
User: Find open issues labeled "bug" in microsoft/vscode
Bot: [Searches GitHub] Found 23 open issues with bug label...
| Component | Status | ID |
|---|---|---|
| GitHub MCP Server | β Active | 78665d40-7447-4612-aa03-74534221ecd1 |
| GitHub Assistant Bot | β Published | 1116f5c2-f901-4a98-8868-4cbc7cdf9851 |
| Slack Package | β Installed | [email protected] |
- Create Slack App
- Get Slack Bot Token
- Get Slack Signing Secret
- Create Hexabot Credentials
- Create Hexabot Slack Source
- Configure Slack Event URLs
- Link Workflow to Slack
- Test Integration
Type: Conversational
Status: Published
Trigger: User messages in connected channels
Flow:
- User sends message β Slack β Hexabot
- AI Agent receives message with conversation history
- Agent uses GitHub MCP tools if needed
- Agent generates response
- Response sent back to user
Capabilities:
- Access GitHub repositories
- Search issues and PRs
- Get release information
- Read repository files
- Search GitHub content
- Maintain conversation context (10 messages)
Edit workflow YAML in Hexabot Admin:
system: ="You are a helpful GitHub assistant specialized in [YOUR SPECIALTY]..."You can add other MCP servers:
- Filesystem MCP
- Database MCP
- Web search MCP
- Custom MCP servers
Examples:
- PR Reviewer Bot - Reviews pull requests
- Issue Triager - Automatically labels and routes issues
- Release Manager - Manages releases and changelogs
- Code Search - Deep code searching across repos
# Check if running
ps aux | grep hexabot
# Check logs
hexabot logs
# or
npm run dev# Via browser
curl http://localhost:3000/api/mcp
# Should return 401 (auth required) - means it's runningIn Hexabot Admin:
- Go to Workflows β GitHub Assistant Bot
- Click Runs tab
- View execution logs and debug info
- MCP API requires Bearer token authentication
- Slack webhook will require signature verification
- Credentials stored encrypted in Hexabot
- Use environment variables for secrets
- Enable HTTPS for production
- Rotate tokens regularly
- Limit MCP server permissions
- Monitor API usage
- Follow
SLACK_QUICK_START.md - Test basic conversation
- Test GitHub queries
- Customize AI agent prompts
- Add more workflows for specific use cases
- Configure App Home in Slack
- Add Hexabot menu buttons
- Add more MCP integrations (Jira, Confluence, etc.)
- Create workflow templates
- Set up production deployment
- Add monitoring and analytics
- Train team on bot usage
WORKFLOW_SUMMARY.md- Workflow architectureSLACK_SETUP_GUIDE.md- Complete Slack guideSLACK_QUICK_START.md- Quick startsetup-slack.sh- Setup helper script
- Slack Channel:
node_modules/hexabot-channel-slack/README.md - Hexabot API:
node_modules/@hexabot-ai/api/
"Workflow not receiving messages" β Check channel source is linked to workflow
"GitHub tools not working" β Verify GitHub MCP server is enabled and running
"Slack URL verification fails" β Check signing secret in Hexabot credentials
"Bot not responding in Slack" β Ensure bot events are subscribed and bot is in channel
- Check documentation files in this directory
- Hexabot Discord community
- GitHub issues for package-specific problems
Current Status:
- Hexabot installed and running
- GitHub MCP server configured
- GitHub Assistant Bot workflow created and published
- Slack package installed
- Slack app created
- Slack credentials configured
- Slack source created
- Workflow linked to Slack
- Integration tested
Next Action: Follow SLACK_QUICK_START.md to complete Slack setup!
You'll know everything is working when:
- β Bot responds to direct messages in Slack
- β Bot responds to @mentions in channels
- β Bot can fetch GitHub data when asked
- β Conversation history is maintained
- β App Home tab works in Slack
Ready to connect Slack? Start with: SLACK_QUICK_START.md π