Describe the feature or problem you'd like to solve
The only indicators for rate limits are either when they trigger and block access, or the one-off warnings for the weekly limit at 75% or 90%.
Proposed solution
Provide a method or UI element that describes current rate limit usage for all rate limits. Other agentic AI services that rely on rate limiting for access already do this (e.g., Codex, Claude Code have a series of progress bars describing 5-hour, daily, weekly limits, etc.). This could either be part of the GitHub web UI, where you already have a progress bar tracking remaining requests, part of the existing status-bar UI in CLI, and/or part of a command like /usage (e.g., a line like 5-hour limit: 40% left | Daily limit: 80% left | Weekly limit: 33% left). If there needs to be a differentiation between global Copilot limits and service-specific ones, then you could also add that as a separate line or as a note with model info.
Surprising users with messages that they have to stop for an hour is annoying and intrusive, especially if it hits while the agent is working mid-turn. Having a lead-up for every limit, not just the weekly, helps users expect and anticipate it when they will hit, which allows them to pivot to another service to work, or pivot in a break organically.
I'm not asking for rate limits to go away -- I understand why that would be an unreasonable request -- but users should have enough information to be able to work around them without just hitting a sudden wall in a way that, from their perspective, is functionally random.
Example prompts or workflows
/usage
* Changes +12345 -12345
Requests 25 Premium (12h 34m 56s)
Tokens ^ 60.0m * v 400.0k * 59.0m (cached) * 250.0k (reasoning)
- Daily limit: 40% remaining (resets in 12h 34m 56s)
- Weekly limit: 80% remaining (resets in 5d 12h 34m 56s)
- All models currently available
Additional context
No response
Describe the feature or problem you'd like to solve
The only indicators for rate limits are either when they trigger and block access, or the one-off warnings for the weekly limit at 75% or 90%.
Proposed solution
Provide a method or UI element that describes current rate limit usage for all rate limits. Other agentic AI services that rely on rate limiting for access already do this (e.g., Codex, Claude Code have a series of progress bars describing 5-hour, daily, weekly limits, etc.). This could either be part of the GitHub web UI, where you already have a progress bar tracking remaining requests, part of the existing status-bar UI in CLI, and/or part of a command like /usage (e.g., a line like
5-hour limit: 40% left | Daily limit: 80% left | Weekly limit: 33% left). If there needs to be a differentiation between global Copilot limits and service-specific ones, then you could also add that as a separate line or as a note with model info.Surprising users with messages that they have to stop for an hour is annoying and intrusive, especially if it hits while the agent is working mid-turn. Having a lead-up for every limit, not just the weekly, helps users expect and anticipate it when they will hit, which allows them to pivot to another service to work, or pivot in a break organically.
I'm not asking for rate limits to go away -- I understand why that would be an unreasonable request -- but users should have enough information to be able to work around them without just hitting a sudden wall in a way that, from their perspective, is functionally random.
Example prompts or workflows
Additional context
No response