Add instructions for LLM agents

[Zoxc]

This adds instructions for LLM agents to help comply with the upcoming rust-lang/rust-forge#1040.

These instructions are intentionally not written with explanations or justifications for humans as that could end up confusing LLMs. They are also written to be minimal again to avoid confusing LLMs.

Unfortunately Claude Code ignores the standard AGENTS.md so CLAUDE.md is added to include it. I tested that Claude Code does mark such comments with [LLM-generated] with these changes.

Agents are asked to mark non-trivial documentation / comments with [LLM-generated] as these are explicitly not allowed by the policy, to help avoid accidentally leaving these in PRs.

It also instructs them to avoid the trap of using cargo directly, as this repo looks like a cargo project and you need to dig into documentation to find out otherwise.

[rustbot]

r? @Mark-Simulacrum

rustbot has assigned @Mark-Simulacrum.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Fallback group: @Mark-Simulacrum, @jieyouxu
• @Mark-Simulacrum, @jieyouxu expanded to Mark-Simulacrum, jieyouxu

[jyn514]

@Zoxc I do not think we should merge this PR. I have been carefully coordinating the communication strategy for the new policy; this is a sensitive topic and this PR is going to affect how people perceive it.