Automations let you run coding agents in the background—on a schedule or triggered by events. Build intelligent workflows that fix bugs, maintain documentation, clear technical debt, and keep your team informed, all without lifting a finger.
What are Automations?
Automations are the best PRs: the ones that come without you having to think about them.
Instead of manually assigning tasks, automations proactively detect issues and take action. A Sentry error gets fixed within seconds of detection. Documentation stays in sync with your code. Security vulnerabilities get caught and patched before they become problems.
Each automation runs in a secure sandbox environment with full repository access and your configured MCP servers.
Triggers
Automations can run on a schedule or react to events from your integrations.
Scheduled Triggers
Run automations at regular intervals:
| Schedule | Use Case |
|---|
| Hourly | Monitor for CI failures, new Sentry errors |
| Daily | Generate changelogs, post standup summaries |
| Weekly | Clean up stale TODOs, security scans, metrics reports |
| Monthly | Dependency audits, documentation reviews |
Event Triggers
React to events from your connected integrations:
| Integration | Events |
|---|
| GitHub | PR opened, PR merged, issue created, comment added |
| GitLab | Merge request opened, merged, issue created |
| Sentry | New error, error spike, regression detected |
| Linear | Issue created, issue updated, status changed |
| Slack | Message in channel, @tembo mention |
Event triggers respond in real-time. A Sentry error can have a fix PR opened within seconds of detection.
Creating an Automation
From a Template
The fastest way to get started:
- Navigate to Automations in your Tembo dashboard
- Click the Templates tab
- Browse available templates and click Use template
- Configure which integration to use (if multiple are connected)
- Customize the instructions if needed
- Enable the automation
From Scratch
For custom workflows:
- Click New Automation
- Give it a descriptive name
- Write your instructions using the rich text editor
- Add triggers (schedule and/or event-based)
- Select which MCP servers to enable
- Choose your preferred coding agent
- Save and enable
Configuration Options
Instructions
Write clear, specific instructions for what the automation should do. You can:
- Use @mentions to reference integrations (e.g., @github, @slack, @linear)
- Include formatting with headers, lists, and code blocks
- Reference specific repositories, channels, or projects
See the prompting guide for best practices.
Triggers
Add one or more triggers to control when the automation runs:
Schedule triggers use cron expressions. Common options:
- Every hour:
0 * * * *
- Daily at 8 AM:
0 8 * * *
- Weekly on Monday:
0 9 * * 1
- Monthly on the 1st:
0 9 1 * *
Event triggers connect to your integrations and fire when specific events occur. You can filter events by repository, project, labels, and more.
MCP Servers
Enable the MCP servers your automation needs. Common combinations:
- Code changes: GitHub/GitLab
- Bug fixes: GitHub + Sentry
- Issue triage: GitHub + Linear
- Team updates: GitHub + Slack
- Full workflow: GitHub + Linear + Slack + Sentry
Agent Selection
Choose which coding agent runs your automation. Different agents have different strengths—Claude Code excels at complex reasoning, while others may be faster for simpler tasks.
Example Automations
Auto-Generate PR Descriptions
Automatically create clear, well-formatted PR descriptions when PRs are opened.
Trigger: PR opened
MCP Servers: GitHub
When a PR is opened, analyze the diff and generate a description:
1. Summarize what changed and why
2. List key modifications
3. Note any breaking changes
4. Add testing instructions
Watch the tutorial →
Prioritize Sentry Errors
Get a daily report of the most impactful bugs, ranked by users affected.
Trigger: Daily at 8 AM
MCP Servers: Sentry, Slack, Linear
Every morning:
1. Fetch errors from the last 24 hours
2. Rank by number of affected users
3. Post top 3 to #engineering-bugs
4. Create Linear tickets for each
Watch the tutorial →
Automated PR Reviews
Get instant code reviews on every pull request.
Trigger: PR opened/updated
MCP Servers: GitHub
Review this PR for:
- Bugs and logic errors
- Security vulnerabilities
- Performance issues
- Style guide compliance
Leave inline comments on specific lines.
Auto-approve if no issues found.
Watch the tutorial →
Security Vulnerability Scan
Scan your codebase daily for security issues.
Trigger: Daily
MCP Servers: GitHub, Linear
Scan for:
- SQL injection
- XSS vulnerabilities
- Hardcoded secrets
- Authentication flaws
Create Linear tickets for each finding with severity ratings.
Generate fix PRs for critical issues.
Watch the tutorial →
Enrich Linear Issues
Automatically add context to new Linear issues from your codebase, Slack, and docs.
Trigger: Linear issue created
MCP Servers: Linear, GitHub, Slack, Notion
When a new issue is created:
1. Search codebase for relevant files
2. Find related Slack conversations
3. Pull in Notion documentation
4. Add a comment with implementation hints
Watch the tutorial →
Managing Automations
Run History
Each automation tracks its run history. View:
- When it last ran
- Success/failure status
- PRs created
- Messages posted
Pausing and Archiving
- Pause an automation to temporarily stop it from running
- Archive automations you no longer need (can be restored later)
Editing Live Automations
Changes to automations take effect on the next run. Running automations complete with their original instructions.
Best Practices
Start Simple
Begin with one automation that solves a clear problem. Validate it works well before adding more complexity or creating additional automations.
Use Templates
Templates are battle-tested starting points. Customize them to fit your workflow rather than building from scratch.
Be Specific
Vague instructions lead to unpredictable results. Tell the automation exactly what to do, where to post, and how to format output.
Test First
Use a test repository or Slack channel when setting up new automations. Once validated, switch to production targets.
Monitor Results
Check run history regularly. If an automation isn’t producing useful results, adjust the instructions.