Ori’s real power emerges when you combine its features: scheduled tasks + connectors + computer use + memory = powerful workflow automation.
Pattern: Morning briefing
Every weekday at 8:30am:
1. Check my Google Calendar for today's meetings
2. Check GitHub for any PRs assigned to me
3. Search for top AI news
4. Give me a 1-minute briefing summary
Pattern: PR review assistant
Every 2 hours during work days, check if there are new PRs
on the ori repo that I haven't reviewed. If there are any,
summarize the changes and flag potential issues.
Pattern: Content monitoring
Every day at noon, check these competitor websites for pricing changes:
- competitor-a.com/pricing
- competitor-b.com/pricing
- competitor-c.com/pricing
Compare with yesterday's prices and alert me if anything changed.
Pattern: Email digest
Every evening at 6pm, search my Gmail for emails I haven't responded to
today that seem important. Give me a prioritized list with suggested responses.
Building complex automations
You can build sophisticated automations by telling Ori your goal in plain language. It will figure out the right combination of tools:
I want a weekly report every Friday at 4pm that includes:
- Lines of code changed this week (from git)
- Number of PRs merged (from GitHub)
- Any open bugs labeled "critical" (from GitHub issues)
- Team velocity compared to last week
Format it as a clean markdown report.
Tips
Start simple, then iterate. Create a basic scheduled task first. Once it’s working, refine the prompt to add more detail or combine more data sources.
Use memory to your advantage. Once you tell Ori your team members, project names, and preferences, scheduled tasks automatically use that context.
Combine with computer use. “Every morning, open Jira in my browser and screenshot my sprint board” — Ori can automate visual workflows too.
