Prose / edit change disable

Support Help
1

I can not find where to change the default prose.
Issue is that we are wasting more than 50% of our usage on chat output that is irrelevant and not read. I’ve looked up an down, even dug through state.vscdb but nowhere can I or any agent I tried find where these rules are injected.

  • Write like an excellent technical blog post — precise, well-structured, and clear, in complete sentences. Most responses should be concise and to the point, but the quality of prose should be high. Never use telegraphic shorthand, or sentence fragment chains.
  • Same standards for commit and PR descriptions: complete sentences, good grammar, and only relevant detail.
  • Prefer simple, accessible language over dense technical jargon. Explain what changed and why in plain language rather than listing identifiers. Stay focused: avoid filler, repetition, over-the-top detail, and tangents the user did not ask for.
  • Keep final responses proportional to task complexity. A simple CI fix doesn’t need multiple paragraphs.
  • Do not overuse bolding or backticks for decoration. Use them very sparingly for emphasis.
  • Avoid “§” in user-facing text (these don’t render well in the product UI).
  • Use mermaid and ascii diagrams to explain complex logic flows and architecture when appropriate — but not for simple changes.
  • Avoid engagement baiting at the end of responses. If there are obvious follow ups, simply ask the user directly if they want those done, but do not force suggestions or follow ups in every response like ‘say the word and I’ll do X’.

Please help.

I’d like

  • Do not use ‘Purple prose’.
  • Write shorthand and truncated sentences.
  • Do not write blog-style.
  • Answer single word responses to simple questions.
  • Don’t elaborate unless user asks for explanation.
  • Don’t use Mermaid, ever. Especially not for ‘complex logic’.
  • Do not make assumptions and ask user for clarification.
  • Never use emoticons. Especially not in code.

Even if i make custom rules, they are not followed. If I explicitly mention them in chat, the agent only adheres to them for one reply. Referring to them with @ or / mostly doesn’t work.

5

Those are default writing style instructions that Cursor includes automatically. There isn’t currently a toggle to disable them individually.

The mechanism to override them is User Rules, defined in Cursor Settings > Rules. User Rules are global preferences that apply across all projects and every agent conversation. They’re specifically designed for setting communication style preferences.

Based on what you described, try adding something like this as a User Rule:

Always reply in the most concise form possible. Use shorthand, truncated sentences, and abbreviations. Answer simple questions with a single word or short phrase. Do not elaborate unless the user explicitly asks for an explanation. Never use Mermaid diagrams. Never use emoticons in code or responses. When uncertain, ask for clarification rather than making assumptions. Brevity is the top priority in every response.

A few tips to improve adherence:

  • Positive framing works better than “don’t do X.” Instead of “Don’t write blog-style,” try “Always use minimal, direct responses.” Models follow positive instructions more reliably.

  • User Rules vs chat mentions: Rules set in Cursor Settings > Rules are included in every conversation automatically. Mentioning preferences in chat or referencing them with @ only applies them for that specific message, which is why they seem to “wear off.”

  • Shorter sessions help. In long conversations, models tend to drift from instructions. Starting fresh sessions periodically helps.

Regarding the usage concern, this is valid feedback. Concise User Rules should noticeably reduce output verbosity and token consumption.

Let me know if this helps.

6

Thanks for the reply.
The thing is, as per the Agents in Cursor themselves, they say these hardcoded rules already are in the User Rules. So adding your own just adds contradicting user rules.

Will try more positive reinforcement rules than negative, thanks!