Product · Accounts & capacity
Capacity is part of the plan
Agents run on your own accounts, and those accounts have limits. Catchscript keeps every account, its usage windows, and its reset time in one place, and uses them to decide when a queued card can start.
Beyond a usage counter
A percentage says how much you used. The factory also needs to know which account, which window, and when it resets.
- 01
Keep account identities together
Review personal and work accounts by provider instead of keeping a mental list of logins.
- 02
Check allowance and reset time
Inspect each account's reported quota windows and reset information. A reset time is not a billing renewal date, and missing data is not proof that an account is ready.
- 03
Schedule agents against capacity
Queued cards wait for both a free slot and an account with room. You choose which accounts the factory may use.
- 04
Stay in control
Account selection is user-directed by default. Automatic failover is opt-in and depends on the integration. Uninterrupted sessions are not promised.
What the current build covers
Development scope. Final compatibility will be confirmed with the public release.
Claude accounts
Account details and provider-reported usage windows for Claude Code. This is not Claude conversation analytics.
OpenAI Codex accounts
Account details, quota windows, and reset information for Codex. ChatGPT conversation limits and API invoices are separate.
Token expiry
Expiry information where the integration exposes it. Token expiry does not tell you when a subscription renews.
Platform
macOS. Catchscript has not launched, and no public download is available yet.
Questions about accounts
Does Catchscript increase my limits?
No. Provider limits still apply. Only connect accounts you are authorized to use, and follow your provider's and organization's policies.
Does it track ChatGPT usage or API costs?
No. The focus is Claude Code and Codex allowance. Codex access through a ChatGPT plan is different from ChatGPT chat limits and from separately billed API usage.
Will it warn before every limit?
Universal advance alerts and exhaustion forecasts are not promised. Available fields depend on each provider.
Where can I check provider rules?
Read Claude's usage-limit documentation and OpenAI's Codex plan documentation. Provider policies can change.
Run more agents without guessing
Capacity-aware scheduling ships with early access.