For things the agent forgets to obey often, at least in Claude Code, there are also "output styles" that are more deeply embedded - into a system prompt - and agent is also periodically reminded of them during the session: https://code.claude.com/docs/en/output-styles

I haven't used them so far but maybe these would work better than basic instructions for such cases.

In practice output styles are just more context (equivalent to CLAUDE.md or memory) ... but with a slightly increased weight.

didn't know of that. I'll switch it to learning mode at work, wondering how long it's going to last.