News for hackers. Stuff that matters.

A comment on Show HN: Claude-thermos – keeps your Claude session warm for you. Parent comment

Re:Show HN: Claude-thermos – keeps your Claude session warm ... by SwellJoe on Thursday July 23, 2026 @05:33PM (#49028354)
Keeping a conversation with a very large model active requires hundreds of GB of memory. If my conversation can never be swapped out, like when I go to lunch or take an afternoon walk, that's several hundred GB not available for other users. If everybody does that, Anthropic needs even more infrastructure than the quarter trillion dollars in infra they're already using.
[ Parent | on HN ]
Re:Show HN: Claude-thermos – keeps your Claude session warm ... by s0ck_r4w on Thursday July 23, 2026 @05:41PM (#49028424)
The idea here is that your conversation is being swapped out for no good reason. The session is still active, as the subagents continue to make progress.

How is this comparable to going to lunch or taking a walk?

[ Parent | on HN ]
Re:Show HN: Claude-thermos – keeps your Claude session warm ... by SwellJoe on Thursday July 23, 2026 @05:48PM (#49028495)
Is that actually happening, though?
[ Parent | on HN ]
Re:Show HN: Claude-thermos – keeps your Claude session warm ... by devnonymous on Thursday July 23, 2026 @05:59PM (#49028618)
Yes https://github.com/izeigerman/claude-thermos/blob/main/src%2...
[ Parent | on HN ]
Re:Show HN: Claude-thermos – keeps your Claude session warm ... by idonotknowwhy on Thursday July 23, 2026 @07:24PM (#49029437)
CC actually prompts Claude about this by default in the ~20k system prompt and instructs it to avoid 300s timeout and to be mindful of the 300s cache expiration.
[ Parent | on HN ]
Re:Show HN: Claude-thermos – keeps your Claude session warm ... by dannyw on Thursday July 23, 2026 @10:19PM (#49030553)
CC also will also block sleeps longer than eg 300s; the harness handholds the model quite a bit.
[ Parent | on HN ]
Re:Show HN: Claude-thermos – keeps your Claude session warm ... by dannyw on Thursday July 23, 2026 @09:55PM (#49030425)
But you pay for cache writes, it's not like you're abusing anything, you're simply choosing to call a billed API for the explicitly intended purpose of writing/refreshing cache.

If you keep this running for hours without doing anything, it will drain your limits and API. The use case of keeping the main thread cache warm while subagents work is very genuine and legitimate.

[ Parent | on HN ]