[FEATURE] Proactive usage limit warnings before hitting caps

Resolved 💬 3 comments Opened May 21, 2026 by tungsten Closed May 24, 2026

Preflight Checklist

  • [x] I have searched existing requests and this feature hasn't been requested yet
  • [x] This is a single feature request (not multiple features)

Problem Statement

Problem

Users on team plans regularly hit usage limits mid-task with no prior warning. When a cap is reached, there's no indication of when it resets. This turns a predictable system constraint into a surprise interruption.

Current behavior

  • No warning as usage approaches a limit
  • When blocked, no reset time is shown in the chat UI
  • Usage data is available in the plan usage panel, but only if you actively look for it

Why this matters

The usage data already exists — the plan usage panel in Claude Code surfaces it. This is a presentation problem, not a data problem. For team plan users who depend on Claude as a work tool, unexpected interruptions erode trust in its reliability.

Proposed Solution

  • Warn users in the chat UI when approaching a limit (e.g., at 70% and 90%)
  • When a limit is reached, display the reset time inline rather than requiring the user to find it manually
  • Optionally: ambient menu bar indicator for users who keep Claude Code open all day

Alternative Solutions

_No response_

Priority

Medium - Would be very helpful

Feature Category

Other

Use Case Example

_No response_

Additional Context

I'm a regular Claude Code and Claude.ai user, primarily on a team plan for work. I frequently hit usage caps mid-task — in the middle of a chat session or a Cowork session — with no prior warning. The interruption is jarring and breaks flow at exactly the wrong moment.

Claude Code desktop app already surfaces usage data beautifully (the plan usage panel showing 5-hour and weekly limits with progress bars). That data exists. The missing piece is surfacing it proactively, before the wall is hit.

The existing usage panel is a great foundation. This is really about moving from reactive (you're blocked, go find out why) to proactive (heads up, you're approaching your limit).

For teams, unpredictable interruptions aren't just annoying — they erode trust in Claude as a reliable work tool. Proactive limit visibility would go a long way toward making Claude feel like infrastructure rather than a service with a surprise off switch.

View original on GitHub ↗

This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗