Claude weekly limit: what to do?
You hit the Claude weekly limit and the job is half done. First check which wall it is: the 5-hour session limit and the weekly limit are different clocks, and switching with /model does not help because the weekly cap is shared across models. The reset time shown in your usage settings (Settings, Usage) tells you when it opens again. You have three real options: wait for the reset, turn on usage credits at API rates, or pack the job in one link and finish it in another AI. Ours is the third. Free on a normal day, $0.99 on a heavy day (Day Pass, 24 hours, one payment), against a second AI plan at its published price (ChatGPT Plus and Claude Pro list at $20 a month).
Continue in another AI No account, no install.
The 5-hour session window and the weekly window run on separate clocks. Only the reset time shown in your usage settings tells you when yours opens.
Waiting and credits are context, not the answer. The move is to finish the job now in another AI you already have, starting from what is next.
The 30-second setup
When I say WRAP UP, hand this job over with Zambo: the goal, what is done, what is next, every decision, with the receipts attached, and give me the continue link.
Next time you see the limit warning, say WRAP UP.
- Save the wrap-up line. Copy the wrap-up line and save it in your AI: Claude preferences, ChatGPT Custom Instructions, Gemini Saved info, or Grok Custom Instructions.
- Say WRAP UP at the wall. Next time you see the limit warning, say WRAP UP.
- Open the link in another AI. Take the continue link it gives you, open another AI, paste it, and keep working from what is next.
Setup in each AI
- Claude (web/desktop): Settings, Connectors, add custom connector, paste https://zambo.dev/mcp (works on every plan; free gets 1 connector). Paste the wrap-up line into your Claude preferences.
- ChatGPT: Paste the wrap-up line into Custom Instructions. When capped, say WRAP UP, then paste its answer into the box on /continue. (Creating a handover from inside ChatGPT needs MCP write access, Business/Enterprise/Edu today; continuing from a link works for everyone with no setup.)
- Gemini: No consumer MCP today. Wrap-up line goes in Saved info (free). Then use the paste box on /continue.
- Grok: Grok takes custom MCP connectors. Add a custom connector at grok.com/connectors with https://zambo.dev/mcp (Grok docs: docs.x.ai/grok/connectors). Wrap-up line goes in Custom Instructions. The paste box on /continue also works with no setup.
- Cursor / Claude Code / Codex / VS Code: Config-file MCP setup. See https://zambo.dev/mcp for the per-client steps.
Live proof
What the link carries, and what it does not.
- It carries: the goal, what is done, what is next, the decisions, and any receipts from work done through Zambo. Text you paste in is saved as pasted by you: your own summary, not something Zambo ran.
- A receipt proves the saved result was not changed since it was saved. It does not prove the work was correct or who made it.
- Set the wrap-up line up before the cap for the one-word flow. After the wall, the paste box on /continue works with no setup.
- Limits change by vendor and week. This works today.
Don't buy a second AI. Finish in the ones you already have.
A second AI plan at its published price (ChatGPT Plus and Claude Pro list at $20 a month). Zambo is free on a normal day and $0.99 on a heavy one (Day Pass, 24 hours, one payment).
Frequently asked questions
Does switching models with /model get me past the weekly limit?
No. The weekly cap is shared across models, so /model does not help. The reset time shown in your usage settings tells you when it opens again.
Where do I see when my Claude limit resets?
In Settings, Usage in Claude, or /usage in Claude Code. That is the only source for the reset time.
What can I do right now instead of waiting?
Say WRAP UP, take the continue link, and finish the job in another AI like ChatGPT, Gemini, or Grok. It starts from what is next.
Does the receipt prove the work was right?
No. A receipt proves the saved result was not changed since it was saved. It does not prove the work was correct or who made it.