What a system prompt does, and which of Anthropic's prompts are published

A GitHub issue opened on 30 October 2026 on the anthropics/claude-code repository says Claude Code began receiving a server-side instruction around October 27 that causes sessions to refuse game-related development work.

Around October 27, Claude Code began receiving a server-side instruction that causes sessions to refuse game-related development work. This is blocking legitimate professional use.

The issue's author asks Anthropic to document server-delivered instruction changes of this scope in the Claude Code changelog.

3. Document server-delivered instruction changes of this scope in the Claude Code changelog, so paying professional users are not blindsided mid project.

The issue does not quote the text of the instruction.

Sessions that performed this work for months now refuse to continue it, citing the new instruction against reproducing games.

Anthropic's prompting guide gives a Messages API example in which a call passes system="You are a helpful coding assistant specializing in Python." as one parameter and a user message asking how to sort a list of dictionaries by key in a separate messages list.

system = "You are a helpful coding assistant specializing in Python." , messages = [ { "role" : "user" , "content" : "How do I sort a list of dictionaries by key?" } ]

Anthropic's prompting guide says setting a role in the system prompt focuses Claude's behavior and tone for your use case, and that even a single sentence makes a difference.

Setting a role in the system prompt focuses Claude's behavior and tone for your use case. Even a single sentence makes a difference:

Anthropic's prompting guide says Claude Opus 4.5 and Claude Opus 4.6 are more responsive to the system prompt than previous models.

Claude Opus 4.5 and Claude Opus 4.6 are also more responsive to the system prompt than previous models.

Anthropic's prompting guide says prompts designed to reduce undertriggering on tools or skills might now overtrigger on these models, and that the fix is to soften aggressive language, for example replacing "CRITICAL: You MUST use this tool when..." with "Use this tool when...".

If your prompts were designed to reduce undertriggering on tools or skills, these models might now overtrigger. The fix is to soften any aggressive language. Where you might have said "CRITICAL: You MUST use this tool when...," you can use more normal prompting such as "Use this tool when..." instead.

Anthropic's system prompts page says Claude's web interface (claude.ai) and mobile apps use a system prompt to provide up-to-date information, such as the current date, to Claude at the start of every conversation.

Claude's web interface (claude.ai) and mobile apps use a system prompt to provide up-to-date information, such as the current date, to Claude at the start of every conversation.

Anthropic's system prompts page says the system prompt also encourages certain behaviors, such as always providing code snippets in Markdown.

The system prompt also encourages certain behaviors, such as always providing code snippets in Markdown.

Anthropic's system prompts page says this prompt is periodically updated to improve Claude's responses, and that some models have multiple dated entries on their pages.

This prompt is periodically updated to improve Claude's responses. ... Some models have multiple dated entries on their pages.

Anthropic's system prompts page says these system prompt updates do not apply to the Claude API.

These system prompt updates do not apply to the Claude API.

Anthropic's Claude Code documentation says Claude Code's system prompt isn't published.

Claude Code's system prompt isn't published.

Anthropic's Claude Code documentation says that to give Claude standing instructions, users can use CLAUDE.md files or the --append-system-prompt flag.

To give Claude standing instructions, use CLAUDE.md files or the --append-system-prompt flag.