For teams whose plan minutes run out before the month ends.
Your setup
- Scheduled runs, runs after each deploy, and chat work with a Machine.
- The minutes go faster than you expected.
- You want the same checks for fewer minutes.
Say this to Certyn
We used most of our monthly minutes in two weeks. What is using them, and how do we use less?
What Certyn does
- Points you to Settings → Usage: minutes used and left, and when they reset. The chat cannot read these numbers.
- Explains what counts:
- A run: its time from start to end. Deep runs count 3 times, Expert 5 times. Stopped and failed runs do not count.
- A Machine: its time while it is on. A cloud Machine stops after 30 minutes with no command. On a runner it stays on until you stop it, at most 8 hours.
- A chat answer: its model cost, 1 cent = 1 minute. Deep and Expert answers cost more.
- Lists your recent runs by trigger (schedule, deploy, manual, CI) and depth, and your processes with their schedules.
- Proposes cuts, and asks before each settings change:
- Fewer schedules: daily instead of hourly, only after each deploy, or off.
- Narrower goals: one flow per run, not "test everything".
- A shorter limit: explore runs take
minutesfrom 10 to 240 (default 60). - Test cases for repeated checks: each test-case run has at most 5 minutes.
- Standard depth where Deep or Expert finds nothing more: Project settings → Testing depth.
- Stop the Machine when you are done.
What you get
- A clear picture of what uses your minutes.
- Schedules, run lengths and depths that match what you need.
Limits
- Settings → Usage shows account totals, not per project or per run. Ask Certyn to list long runs.
- Scheduled and deploy-triggered runs of a goal have 60 minutes. You cannot set a shorter limit for them. Write a narrow goal.
minutesis a limit: a run counts only the time it really took.- When the minutes are used up, new runs and Machines do not start, and the chat answers only at Standard. Upgrade in Settings → Billing.