
Claude Usage Tracker 作者: lugia19
Helps you track your claude.ai usage caps based on tokens sent, and syncs across devices using firebase.
366 个用户366 个用户
您需要 Firefox 来使用此扩展
扩展元数据
屏幕截图

关于此扩展
This extension helps you monitor how much of your Claude usage quota remains. It calculates token consumption from various sources including uploaded files, project knowledge, chat history, and AI responses.
Most elements in the chat UI (Namely the length, cost, estimate, caching status) have a tooltip explaining them further! Be sure to check them out.
The extension tracks token usage from:
- Files - Documents uploaded to chats or synced via Google Drive, Github, etc
- Projects - Knowledge files and custom instructions
- Personal preferences - Your configured settings
- Message history - Full conversation context
- System prompts - Enabled tools (analysis, artifacts) on a per-chat basis
- MOST MCPs/Integrations - There are some limitations in cases where a "Knowledge" object is returned that I can't access, such as with web search
Limitations:
- Web search results - The full results are not exposed in the conversation history, so I can't track them properly (See above, Knowledge objects)
- Research - Most of it happens on the backend, so I can't track it at all
Token calculation is handled either through Anthropic's API (if you provide your key) or via gpt-tokenizer.
It will additionally fetch your organization ID on claude.ai to synchronize your usage amounts across devices via firebase, as well as to store data on the amount of tokens you consumed before hitting the usage cap (to crowdsource more accurate counts).
The source code can be found here: https://github.com/lugia19/Claude-Usage-Extension
About the required permissions:
Host Permissions:
- Claude.ai: Needed to read the tokens and usage amounts.
- Github: Needed to fetch the up to date usage caps and settings from the repo.
Actual permissions:
- Tabs: To send updates from the background script to the content scripts
- Storage: To store and track usage amounts
- Alarms: To periodically sync with firebase, and check for the cap reset
- Cookies: To get the sessionKey required to fetch conversation content (This is never stored, not even locally, just used temporarily - blame MV3 for this being required)
- webRequest: To detect when a new message is sent, or data recieved
- notifications: To notify you when your usage has been re-filled
Source code here: https://github.com/lugia19/Claude-Usage-Extension
Most elements in the chat UI (Namely the length, cost, estimate, caching status) have a tooltip explaining them further! Be sure to check them out.
The extension tracks token usage from:
- Files - Documents uploaded to chats or synced via Google Drive, Github, etc
- Projects - Knowledge files and custom instructions
- Personal preferences - Your configured settings
- Message history - Full conversation context
- System prompts - Enabled tools (analysis, artifacts) on a per-chat basis
- MOST MCPs/Integrations - There are some limitations in cases where a "Knowledge" object is returned that I can't access, such as with web search
Limitations:
- Web search results - The full results are not exposed in the conversation history, so I can't track them properly (See above, Knowledge objects)
- Research - Most of it happens on the backend, so I can't track it at all
Token calculation is handled either through Anthropic's API (if you provide your key) or via gpt-tokenizer.
It will additionally fetch your organization ID on claude.ai to synchronize your usage amounts across devices via firebase, as well as to store data on the amount of tokens you consumed before hitting the usage cap (to crowdsource more accurate counts).
The source code can be found here: https://github.com/lugia19/Claude-Usage-Extension
About the required permissions:
Host Permissions:
- Claude.ai: Needed to read the tokens and usage amounts.
- Github: Needed to fetch the up to date usage caps and settings from the repo.
Actual permissions:
- Tabs: To send updates from the background script to the content scripts
- Storage: To store and track usage amounts
- Alarms: To periodically sync with firebase, and check for the cap reset
- Cookies: To get the sessionKey required to fetch conversation content (This is never stored, not even locally, just used temporarily - blame MV3 for this being required)
- webRequest: To detect when a new message is sent, or data recieved
- notifications: To notify you when your usage has been re-filled
Source code here: https://github.com/lugia19/Claude-Usage-Extension
评分 5(1 位用户)
支持这位开发者
此扩展的开发者寻求您的小小捐助以支持其持续发展。
权限与数据详细了解
必要权限:
- 为您显示通知
- 获取浏览器标签页
- 访问您在 claude.ai 的数据
可选权限:
- 访问您在 claude.ai 的数据
- 访问您在 raw.githubusercontent.com 的数据
- 访问您在 github.com 的数据
- 访问您在 api.anthropic.com 的数据
更多信息
- 附加组件链接
- 版本
- 4.0
- 大小
- 1.85 MB
- 上次更新
- 1 天前 (2025年7月23日)
- 相关分类
- 许可证
- 仅 GNU 通用公共许可证 v3.0
- 隐私政策
- 阅读此附加组件的隐私政策
- 版本历史
- 标签
- 添加到收藏集
4.0 的发布说明
Complete refactor of the code.
Added caching support (with indicators).
Added tooltips on hover/long tap to explain things about the extension.
Added caching support (with indicators).
Added tooltips on hover/long tap to explain things about the extension.
lugia19 制作的更多扩展
- 目前尚无评分
- 目前尚无评分
- 目前尚无评分
- 目前尚无评分
- 目前尚无评分
- 目前尚无评分