# How would you make the most of practically unlimited tokens for a few days? — retained record

> Retained reddit record from reddit:ClaudeCode.

Canonical: https://highsignal.app/data/records/d3d43077594dabfd

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:ClaudeCode)  published 2026-10-20 10:01 ingested 2026-10-21 02:40

[original source ↗](https://www.reddit.com/r/ClaudeCode/comments/1xao2sp/how_would_you_make_the_most_of_practically/)

## retained text



```
I keep reading about people building surprisingly large projects in days or weeks with AI agents, and I'm wondering how much of that comes down to their workflow.

I'm currently working on several projects with predefined tasks and dependencies. My usual approach is fairly sequential: pick a task whose dependencies are satisfied, hand it to an agent, let it drive the implementation to completion, review the result, then move on to the next task.

Now I have a few days with practically unlimited token usage before going back to my regular Max x20 allowance, and I want to make the most of it.

Would you go all-in on Fable for everything? Ultracode? Run multiple agents in parallel? Use separate agents for implementation, testing, review, and architecture? Or something completely different?

The goal is to push several real projects as far as possible, but without sacrificing quality. I don't want to spend the following weeks cleaning up a mountain of AI-generated garbage.

For those who've managed to make massive progress with agents in a short time, what actually worked? What was your setup and workflow, and what would you do differently if token usage wasn't a constraint?
```



document fetched 2026-10-21 00:40 · canonical https://reddit.com/r/ClaudeCode/comments/1xao2sp/how_would_you_make_the_most_of_practically
