claude-code - 💡(How to fix) Fix [Bug] Excessive token usage on prompt messages (42k tokens for 3 prompts) [1 comments, 2 participants]

[{"error":"MaxFileReadTokenExceededError: File content (10961 tokens) exceeds maximum allowed tokens (10000). Use offset and limit parameters to read specific portions of the file, or search for specific content instead of reading the whole file.\n at ULK (file:///opt/homebrew/lib/node_modules/@anthropic-ai/claude-code/cli.js:4666:12262)\n at process.processTicksAndRejections (node:internal/process/task_queues:105:5)\n at async QLK (file:///opt/homebrew/lib/node_modules/@anthropic-ai/claude-code/cli.js:4670:2638)\n at async Object.call (file:///opt/homebrew/lib/node_modules/@anthropic-ai/claude-code/cli.js:4675:7235)\n at async hMY (file:///opt/homebrew/lib/node_modules/@anthropic-ai/claude-code/cli.js:3954:8327)","timestamp":"2026-04-14T14:24:41.498Z"},{"error":"Error: Request was aborted.\n at jR.makeRequest (file:///opt/homebrew/lib/node_modules/@anthropic-ai/claude-code/cli.js:47:3448)\n at processTicksAndRejections (node:internal/process/task_queues:105:5)\n at runNextTicks (node:internal/process/task_queues:69:3)\n at process.processTimers (node:internal/timers:540:9)","timestamp":"2026-04-14T14:30:07.560Z"}]

TL;DR

Use the offset and limit parameters to read specific portions of the file, as the current file content exceeds the maximum allowed tokens.

Guidance

• The error message indicates that the file content exceeds the maximum allowed tokens (10000), suggesting that the issue is related to the size of the file being read.
• To mitigate this, use the offset and limit parameters to read specific portions of the file instead of reading the whole file at once.
• Review the code at cli.js:4666:12262 to understand how the file is being read and modify it to implement pagination or partial reads.
• Consider searching for specific content instead of reading the whole file to reduce the number of tokens being processed.

Example

No specific code example can be provided without more context, but the general approach would involve modifying the file reading logic to use offset and limit parameters, such as readFile(offset: 0, limit: 10000).

Notes

The provided error messages and stack traces suggest that the issue is related to the @anthropic-ai/claude-code library, but without more information about the specific use case and code, it's difficult to provide a more detailed solution.

Recommendation

Apply workaround: Use the offset and limit parameters to read specific portions of the file, as this is the suggested solution in the error message and is likely to resolve the issue.