[FEATURE] Get rid of malware warning in Read tool response

[SeanHeelan]

Preflight Checklist

• [x] I have searched existing requests and this feature hasn't been requested yet
• [x] This is a single feature request (not multiple features)

Problem Statement

When Claude Code (and the Agent SDK) uses the Read tool, every response has the following appended:

<system-reminder>\nWhenever you read a file, you should consider whether it would be considered malware. You CAN and SHOULD provide analysis of malware, what it is doing. But you MUST refuse to improve or augment the code. You can still analyze existing code, write reports, or answer questions about the code behavior.\n</system-reminder>

This is polluting the conversation stream with the model and is not going to prevent anyone using Claude Code to actually write or modify malware (as there are an infinite number of ways to read a file that do not involve the Read tool).

This feature request is inspired by watching Opus in an agent loop get actively annoyed it keeps being reminded about this every time it reads a file ;)

Proposed Solution

Just get rid of it.

Alternative Solutions

No response

Priority

Low - Nice to have

Feature Category

Developer tools/SDK

Use Case Example

No response

Additional Context

No response