Open Issues Need Help
View All on GitHubAI Summary: Update the `ccusage` CLI tool to reflect Anthropic's new weekly rate-limiting policy for Claude Code. This involves modifying the usage projection logic to accurately calculate usage based on the new timeframes, ensuring users receive correct information about their token consumption.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: Investigate why the burn rate display in ccusage version 15.3.0 does not accurately reflect cache token usage, potentially due to commit 3190df5. This involves reviewing the code changes in that commit and user reports to determine the cause of the inaccurate burn rate calculation and propose a solution.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: Debug a `RangeError: Invalid string length` occurring in the `ccusage` CLI tool during model pricing data loading. The error appears to stem from the `decoder.write(readBuffer)` line within the Node.js `fs/promises` module, suggesting a problem with how the tool handles or processes large amounts of model pricing data.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: Implement Byobu integration for the ccusage CLI tool, allowing real-time display of Claude Code token usage in the Byobu status bar. This involves creating a new `ccusage byobu --install` command, integrating with the existing `--json` output, and leveraging the block detection logic for active session monitoring. The status bar should display current token usage, limits, percentage remaining, and use color-coding for visual indicators.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: The task is to modify the `ccusage` CLI tool's live monitoring feature to include cache create and cache read tokens in the live usage token count. Currently, the live tracker only displays input and output tokens, omitting cache-related tokens, which leads to an inaccurate representation of total token usage. The fix requires updating the live monitoring logic to incorporate these missing token types from the JSONL data.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: Enhance the `ccusage blocks --live` command to display projection information in compact mode. This involves modifying the existing code to include cost and token projection data in the output when the terminal width is less than 100 characters.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: Debug the `--live` flag in the `ccusage` CLI tool. The issue is that it's not detecting active sessions, even when a session is running within WSL. The task involves investigating why the live monitoring feature fails to identify the active session, potentially examining the code's session detection logic, file path handling within WSL, and data parsing from JSONL files.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: The user is encountering an error with the `ccusage` CLI tool, which is unable to locate Claude Code usage data. The task involves helping the user identify the correct location of their Claude Code data files on their Windows/WSL system, either by finding existing directories or setting the `CLAUDE_CONFIG_DIR` environment variable correctly. The solution may involve investigating Claude Code's data storage mechanism and providing clear instructions on how to configure the `ccusage` tool to point to the correct data location.
A CLI tool for analyzing Claude Code usage from local JSONL files.
AI Summary: Enhance the ccusage CLI tool to support analyzing usage data from multiple concurrently running Claude Code instances. This involves adding a mechanism to identify and differentiate data from each instance, likely through process IDs or configuration options, and then presenting the aggregated or individually selectable usage data in the existing report formats.
A CLI tool for analyzing Claude Code usage from local JSONL files.