Issue Station
Sign In
Tools hereToolbox

openclaw - 💡(How to fix) Fix [Bug]: file read multiple times even when context is not full.

Official PRs (…)
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.

Helpful · Quick feedback

Loading…

I am on v2026.5.10-beta.5 but have noticed that versions after 2026.4.23 re-reads files that they already have read and that should be part of the context.

I am not an expert to begin with, but my understanding is that everything the agent does, reads, and does is part of the context it is sending to the LLM provider over and over.

So if an agent reads, let's say, file A.txt, it should be part of the context until there is a compression. But what I have noticed is that in a conversation the agent can read file a.txt multiple times. and not use the one it already in the context.

This is very inefficient. 1st because the agent needs to use tools, so there is back and forth to even start reading the file again. Then the content of the file is again added to the context of the agent. Meaning less room for other context.

Root Cause

This is very inefficient. 1st because the agent needs to use tools, so there is back and forth to even start reading the file again. Then the content of the file is again added to the context of the agent. Meaning less room for other context.

RAW_BUFFERClick to expand / collapse

Bug type

Regression (worked before, now fails)

Beta release blocker

No

Summary

I am on v2026.5.10-beta.5 but have noticed that versions after 2026.4.23 re-reads files that they already have read and that should be part of the context.

I am not an expert to begin with, but my understanding is that everything the agent does, reads, and does is part of the context it is sending to the LLM provider over and over.

So if an agent reads, let's say, file A.txt, it should be part of the context until there is a compression. But what I have noticed is that in a conversation the agent can read file a.txt multiple times. and not use the one it already in the context.

This is very inefficient. 1st because the agent needs to use tools, so there is back and forth to even start reading the file again. Then the content of the file is again added to the context of the agent. Meaning less room for other context.

Steps to reproduce

ask questions about a file, in multiple turns.

Expected behavior

The file gets read 1x and is part of the context, and every other question is answered based on that 1st file read action.

Actual behavior

Agent keeps reading the file with every new question you ask. ok maybe not always but I had an agent do that at least 4 times while the context window was not full.

OpenClaw version

v2026.5.10-beta.5

Operating system

Ubuntu

Install method

npm

Model

Minimax m2.7

Provider / routing chain

minimax/minimax m2.7

Additional provider/model setup details

No response

Logs, screenshots, and evidence

Impact and severity

No response

Additional information

No response

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?
Loading…

FAQ

Expected behavior

The file gets read 1x and is part of the context, and every other question is answered based on that 1st file read action.

#API middleware#SSR setup#ISR setup#authentication setup#request error

Still need to ship something?

×6

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

Back to top recommendations

TRENDING