openclaw - ✅(Solved) Fix fix(agents): context overflow not detected for Bedrock, Ollama, and Cohere providers [2 pull requests, 1 comments, 2 participants]

openclaw2026-04-01 07:51:45

ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

GitHub issue URL

Helpful · Quick feedback

GitHub stats

openclaw/openclaw#58839•Fetched 2026-04-08 02:32:00

View on GitHub

Comments

Participants

Timeline

Reactions

Author

aaron-he-zhu

Participants

aaron-he-zhu

wirjo

Timeline (top)

cross-referenced ×2commented ×1

Error Message

| Provider | Error Message | Detected? |

RAW_BUFFERClick to expand / collapse

Bug Description

isContextOverflowError() and isLikelyContextOverflowError() in src/agents/pi-embedded-helpers/errors.ts detect context window overflow for major providers (Anthropic, OpenAI, Groq) but miss several others. When context overflow goes undetected, the agent retries with the same oversized payload instead of triggering compaction — wasting API credits on requests that will always fail.

Missing Provider Patterns

Provider	Error Message	Detected?
AWS Bedrock	`ValidationException: The input is too long`	No
AWS Bedrock	`ValidationException: Input exceeds the max input token limit`	No
AWS Bedrock	`ModelStreamErrorException: Input is too long`	No
Ollama	`ollama: context length exceeded for llama3`	No
Cohere	`total tokens exceeds the model's max of 4096`	No
Generic	`input is too long for model gpt-5.4`	No
Generic	`maximum number of input tokens`	No

Impact

Users on Bedrock, Ollama, or Cohere hit context limits and see repeated retry failures instead of automatic compaction
Each failed retry wastes input tokens (the full oversized context is sent to the provider)
With 3 retries per attempt, a single undetected overflow wastes 3x the context window in tokens

Expected Behavior

isContextOverflowError() should detect provider-specific context overflow messages so the compaction pipeline can trigger automatically.

Affected Code

src/agents/pi-embedded-helpers/errors.ts:194 — isContextOverflowError()
src/agents/pi-embedded-helpers/errors.ts:248 — isLikelyContextOverflowError()

extent analysis

TL;DR

Update the isContextOverflowError() and isLikelyContextOverflowError() functions in src/agents/pi-embedded-helpers/errors.ts to include patterns for missing providers.

Guidance

Identify and add the missing error message patterns for AWS Bedrock, Ollama, Cohere, and Generic providers to the isContextOverflowError() function.
Update the isLikelyContextOverflowError() function to account for the new patterns and ensure it correctly identifies likely context overflow errors.
Verify the updated functions by testing with the provided error messages and checking if the compaction pipeline triggers automatically.
Consider adding a fallback mechanism to handle unknown error messages and prevent retries with oversized payloads.

Example

// Example of adding a new pattern to isContextOverflowError()
const isContextOverflowError = (error: string) => {
  // Existing patterns...
  const awsBedrockPattern = /ValidationException: The input is too long|ValidationException: Input exceeds the max input token limit|ModelStreamErrorException: Input is too long/;
  const ollamaPattern = /ollama: context length exceeded for llama3/;
  const coherePattern = /total tokens exceeds the model's max of 4096/;
  const genericPattern = /input is too long for model|maximum number of input tokens/;

  return (
    // Existing conditions...
    awsBedrockPattern.test(error) ||
    ollamaPattern.test(error) ||
    coherePattern.test(error) ||
    genericPattern.test(error)
  );
};

Notes

The updated functions should be thoroughly tested to ensure they correctly identify context overflow errors for all providers and trigger the compaction pipeline as expected.

Recommendation

Apply workaround by updating the isContextOverflowError() and isLikelyContextOverflowError() functions to include the missing provider patterns, as this will allow the compaction pipeline to trigger automatically and prevent wasted API credits.

Vote matrix · Quick signals

Works

Did the solution work? Tap to confirm.

Easy Fix

Was it a quick fix?

Time Saver

Did it save you time?

Blocking

Was it severely blocking?

Common Issue

Are others likely hitting this too?

Flaky / Intermittent

Is it intermittent?

Verified / Reproducible

Can you reproduce it reliably?

#api #request error #file not found #serialization error #model compatibility

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

openclaw - ✅(Solved) Fix fix(agents): context overflow not detected for Bedrock, Ollama, and Cohere providers [2 pull requests, 1 comments, 2 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

Fix Action

Fixed

PR fix notes

PR #58847: fix(agents): handle context overflow consistently across providers

Description (problem / solution / changelog)

Summary

Test plan

Changed files

PR #58859: fix(agents): handle context overflow consistently across providers