You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Some of us run more than one Claude Code configuration on the same machine — e.g. a personal account and a work account, each with its own ~/.claude* directory and its own session transcripts. Today this tool reads a single location, so there's no way to see usage and cost across those accounts.
You either point it at one and lose the other, or run separate copies and tally by hand.
The idea
Auto-discover every ~/.claude* directory that contains a projects/ subfolder (e.g. ~/.claude, ~/.claude-perso, ~/.claude-work) and let you view each one's usage, without any configuration !
Sketch of how it could work :
Each discovered directory is treated as an "account" with its own usage DB stored inside it.
The dashboard gets a small tab bar to switch between accounts; each tab loads that account's models, sessions, and charts.
CLI commands (scan, today, stats) iterate over all discovered accounts with a [label] prefix per account
If only one account exists, nothing changes, the tab bar hides itself, so it's fully backward-compatible
Zero config is the key design goal: just name your directories ~/.claude-* and they're picked up automatically. No auth, no integrations, no new product surface, it stays squarely within "parse local JSONLs, show usage and cost," just across N local directories instead of one.
Why I'm posting this here first
I put together a working implementation, but the maintainers (rightly) want feature ideas validated with the community before code lands, to keep the tool tightly scoped and maintainable solo. So:
If this would be useful to you, please 👍 the post or comment with your setup. Specifically I'd love to hear:
Do you actually run multiple ~/.claude* directories? How many, and why (work/personal, multiple orgs, testing)?
Would per-account tabs in the dashboard cover your need, or would you rather see a single merged/aggregate view?
Any naming convention concerns, is keying off the ~/.claude* glob reasonable, or would you want an explicit opt-in list?
If it gets traction, I'm happy to re-submit the PR (Originally #12)
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
The problem
Some of us run more than one Claude Code configuration on the same machine — e.g. a personal account and a work account, each with its own
~/.claude*directory and its own session transcripts. Today this tool reads a single location, so there's no way to see usage and cost across those accounts.You either point it at one and lose the other, or run separate copies and tally by hand.
The idea
Auto-discover every ~/.claude* directory that contains a projects/ subfolder (e.g.
~/.claude,~/.claude-perso,~/.claude-work) and let you view each one's usage, without any configuration !Sketch of how it could work :
Zero config is the key design goal: just name your directories ~/.claude-* and they're picked up automatically. No auth, no integrations, no new product surface, it stays squarely within "parse local JSONLs, show usage and cost," just across N local directories instead of one.
Why I'm posting this here first
I put together a working implementation, but the maintainers (rightly) want feature ideas validated with the community before code lands, to keep the tool tightly scoped and maintainable solo. So:
If this would be useful to you, please 👍 the post or comment with your setup. Specifically I'd love to hear:
~/.claude*directories? How many, and why (work/personal, multiple orgs, testing)?If it gets traction, I'm happy to re-submit the PR (Originally #12)
Beta Was this translation helpful? Give feedback.
All reactions