claude-code - 💡(How to fix) Fix [Bug] Sub-agents deploy successfully but fail to accept prompts at runtime [1 comments, 2 participants]

[{"error":"Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"prompt is too long: 212082 tokens > 200000 maximum"},"request_id":"req_011CajcQ7usVEZyCV5nR9Hrc"}\n at generate (/$bunfs/root/src/entrypoints/cli.js:11:58059)\n at makeRequest (/$bunfs/root/src/entrypoints/cli.js:50:4949)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-05-05T14:52:17.593Z"},{"error":"Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"prompt is too long: 212082 tokens > 200000 maximum"},"request_id":"req_011CajcRMj7HSfPQG3XZxn8D"}\n at generate (/$bunfs/root/src/entrypoints/cli.js:11:58059)\n at makeRequest (/$bunfs/root/src/entrypoints/cli.js:50:4949)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-05-05T14:52:34.296Z"},{"error":"Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"prompt is too long: 212082 tokens > 200000 maximum"},"request_id":"req_011CajcS1C7CjDdPmJXmwGB9"}\n at generate (/$bunfs/root/src/entrypoints/cli.js:11:58059)\n at makeRequest (/$bunfs/root/src/entrypoints/cli.js:50:4949)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-05-05T14:52:43.540Z"}]

TL;DR

The sub-agents fail to work due to the prompt being too long, exceeding the 200000 token maximum.

Guidance

• Verify the prompt length by checking the number of tokens in the request.
• Consider shortening the prompt to be within the 200000 token limit.
• Review the generate function in cli.js to understand how the prompt is being processed and if there are any opportunities to optimize it.
• Check if there are any configuration options to increase the maximum prompt length, if necessary.

Example

No code snippet is provided as the issue is related to the prompt length and not a specific code implementation.

Notes

The errors indicate that the prompt is too long, which suggests that the issue is not with the code itself but rather with the input being provided. The solution will depend on the specific requirements of the application and the constraints of the platform being used.

Recommendation

Apply workaround: Shorten the prompt to be within the 200000 token limit, as this is the most straightforward solution to resolve the issue.