Slack Commands Reference
This reference covers all slash commands and agent @mention patterns available in the Bike4Mind Slack integration.
Slash Commands
/b4m
The primary slash command for Bike4Mind actions.
| Command | Description | Example |
|---|---|---|
/b4m help | Show available commands and agents | /b4m help |
/b4m status | Check your connection status (Slack, GitHub, Atlassian) | /b4m status |
/notebook
Create and manage notebooks directly from Slack.
| Command | Description | Example |
|---|---|---|
/notebook create [title] | Create a new notebook | /notebook create Sprint Retrospective |
/notebook list | List your recent notebooks | /notebook list |
Agent @Mentions
@mention an agent to interact with it. Agents read the full thread context when mentioned in a thread.
@agent — General Assistant
The general-purpose AI agent for questions, explanations, and help.
| Pattern | Description | Example |
|---|---|---|
@agent [question] | Ask any question | @agent What's the difference between REST and GraphQL? |
@agent help | Show all available agents and commands | @agent help |
@agent summarize | Summarize the current thread | @agent summarize this discussion |
@dev — Developer Agent
Specialized for GitHub operations and technical tasks.
| Pattern | Description | Example |
|---|---|---|
@dev create issue [description] | Create a GitHub issue from conversation context | @dev create issue for the login timeout bug |
@dev search [query] | Search code across repositories | @dev search for useAuth hook |
@dev list issues [filters] | List GitHub issues | @dev list open bugs in the frontend repo |
@dev create pr [description] | Create a pull request | @dev create a PR for the feature branch |
@dev show workflow [repo] | Show recent CI/CD workflow runs | @dev show failing workflows on main |
When you @mention @dev in a thread, it analyzes the entire conversation to create well-formed issues with proper titles, descriptions, and labels extracted from the discussion.
@pm — Project Manager Agent
Specialized for Jira operations and project planning.
| Pattern | Description | Example |
|---|---|---|
@pm create ticket [description] | Create a Jira ticket | @pm create ticket for the feature request |
@pm create epic [description] | Create an epic from conversation | @pm create epic from this planning discussion |
@pm search [JQL or natural language] | Search Jira issues | @pm find open bugs in the mobile project |
@pm update [issue-key] | Update a Jira issue | @pm move PROJ-123 to In Progress |
@pm sprint status | Show current sprint status | @pm what's the sprint burndown? |
@analyst — Data Analyst Agent
Specialized for data analysis and insights.
| Pattern | Description | Example |
|---|---|---|
@analyst summarize [topic] | Analyze and summarize data | @analyst summarize our sprint metrics |
@analyst compare [items] | Compare metrics or approaches | @analyst compare Q3 and Q4 velocity |
@analyst insights [data] | Extract insights from shared data | @analyst what patterns do you see in these error rates? |
@researcher — Research Agent
Specialized for documentation and knowledge management.
| Pattern | Description | Example |
|---|---|---|
@researcher find [topic] | Search documentation and knowledge bases | @researcher find our API auth docs |
@researcher create page [title] | Create a Confluence page | @researcher create a Confluence page from this meeting |
@researcher summarize [doc] | Summarize a document or set of results | @researcher summarize our deployment runbook |
Natural Language Support
All agents support natural language variations. You don't need to use exact command syntax:
@dev create an issue for the timeout bug we discussed
@dev Can you make a GitHub issue about the login problem?
@dev I need a bug report for the API rate limiting issue
All three produce similar results — a GitHub issue with context from the conversation.
Agent Response Patterns
Single Response
For simple queries, agents respond with a single message:
You: @agent What is a webhook?
Agent: A webhook is an HTTP callback that sends real-time data...
Threaded Response
For complex operations, agents respond in a thread to keep the channel clean:
You: @dev create issue for the login timeout bug
Agent: (in thread) Created GitHub issue #456: "Login timeout on slow connections"
- Repository: org/frontend
- Labels: bug, authentication
- Link: https://github.com/org/frontend/issues/456
Confirmation Before Action
For destructive or significant actions, agents ask for confirmation:
You: @pm delete PROJ-123
Agent: Are you sure you want to delete PROJ-123 "Update login flow"?
This action cannot be undone. React with ✅ to confirm.
Tips and Best Practices
- Be specific — "create an issue about the login bug" is better than "create an issue"
- Use threads — @mention agents in threads to give them conversation context
- One action per message — agents handle one request at a time
- Check connections — use
/b4m statusto verify your integrations are connected - Right agent for the job — use
@devfor GitHub,@pmfor Jira,@researcherfor Confluence
The Bike4Mind bot must be invited to a channel before agents can respond there. Use /invite @Bike4Mind Bot to add it.
Permissions
| Action | Requires |
|---|---|
| @mention any agent | Bike4Mind account + Slack linked |
| Create GitHub issues | GitHub connected |
| Create Jira tickets | Atlassian connected |
| Create Confluence pages | Atlassian connected |
| Use slash commands | Bot invited to channel |
Rate Limits
Slack rate limits apply to agent responses:
| Scenario | Limit |
|---|---|
| Messages per channel | 1 message/second |
| File uploads | 20 per minute |
| API calls (general) | Varies by method |
Agents automatically handle rate limiting with retries. You may notice brief delays during peak usage.
Related Documentation
- Slack Integration Guide — Overview and getting started
- Slack Admin Guide — App setup and workspace management
- External Integrations Overview — All integrations at a glance