claude-code - 💡(How to fix) Fix [Bug] Subagent spawning fails with xhigh/max reasoning due to prompt length limits [1 comments, 2 participants]

[{"error":"Error: NON-FATAL: Lock acquisition failed for /Users/pz/.local/share/claude/versions/2.1.126 (expected in multi-process scenarios)\n at r86 (/$bunfs/root/src/entrypoints/cli.js:2768:2177)\n at HD8 (/$bunfs/root/src/entrypoints/cli.js:2768:1257)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-05-04T18:55:15.252Z"},{"error":"Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"prompt is too long: 213350 tokens > 200000 maximum"},"request_id":"req_011Cai3T9vQmUgHA5Kf7Jgi1"}\n at generate (/$bunfs/root/src/entrypoints/cli.js:11:57327)\n at makeRequest (/$bunfs/root/src/entrypoints/cli.js:50:4943)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-05-04T18:59:33.796Z"},{"error":"Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"prompt is too long: 213350 tokens > 200000 maximum"},"request_id":"req_011Cai3TxgG6i4uPJigd97K1"}\n at generate (/$bunfs/root/src/entrypoints/cli.js:11:57327)\n at makeRequest (/$bunfs/root/src/entrypoints/cli.js:50:4943)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-05-04T18:59:45.199Z"},{"error":"Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"prompt is too long: 213350 tokens > 200000 maximum"},"request_id":"req_011Cai3UZ3Ut91sdDsf11NQB"}\n at generate (/$bunfs/root/src/entrypoints/cli.js:11:57327)\n at makeRequest (/$bunfs/root/src/entrypoints/cli.js:50:4943)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-05-04T18:59:53.050Z"}]

TL;DR

Reduce the prompt length to below 200,000 tokens to avoid hitting the length limit when spawning subagents.

Guidance

• Identify the specific prompts causing the length limit issue and consider shortening them or splitting the tasks into smaller, more focused prompts.
• Review the auto-save client code, server auto-save handlers, and Sentry config to ensure they are not contributing to the excessive prompt length.
• Investigate if there are any configuration options or parameters that can be adjusted to increase the prompt length limit, although this may not be feasible given the error messages.

Example

No specific code snippet can be provided without more context, but the goal would be to refactor the prompts to be more concise, such as focusing on specific aspects of the auto-save functionality or error handling.

Notes

The issue seems to be directly related to the prompt length exceeding the maximum allowed limit of 200,000 tokens. The provided error messages consistently point to this issue, suggesting that addressing the prompt length will resolve the problem.

Recommendation

Apply workaround: Reduce the prompt length to comply with the 200,000 token limit, as directly addressing the root cause (the limit itself) may not be feasible without further modifications to the system or its configuration.