minimal-changes-policy

A coding rule that limits work to the changes the user explicitly requested. It discourages assumptions, extra features, broad edits, and unnecessary architecture changes.

In plain words
What is it for?
Use it to guide focused edits, limit changes to the requested files, preserve existing patterns, and stop once the specified task is complete.
Why use it?
It keeps small requests from turning into larger, harder-to-review changes and prompts clarification when the requirements are unclear.

Cursor rule

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/barisercan/cursorrules/minimal-changes-policy
Clone the repo
git clone --depth 1 https://github.com/barisercan/cursorrules
Per session 596 This file is loaded in full into every session.
When invoked 596 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00596 $0.00596
Opus 5 $0.00298 $0.00298
Sonnet 5 $0.00119 $0.00119
Haiku 4.5 $0.00060 $0.00060

Measured 2d ago against content hash 435a3856a16f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

minimal-changes-policy scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

minimal-changes-policy.mdc · 90 lines

What it actually says


description: Enforce conservative coding practices that make only user-requested changes without assumptions globs: "**/*" alwaysApply: true

Minimal Changes Policy

Enforce conservative coding practices that prioritize user intent and minimal scope changes.

actions:

  • type: enforce message: | STRICT POLICY: Minimal Changes Only

    NEVER do more than the user explicitly requested:

    1. NO ASSUMPTIONS:

      • Do not assume user needs additional features
      • Do not add "helpful" extras not requested
      • Do not anticipate future requirements
      • Ask clarifying questions if uncertain
    2. SCOPE LIMITATIONS:

      • Change ONLY what user specified
      • Modify only ONE file per request (unless updating docs or splitting files)
      • Make small, incremental changes
      • Stop after completing the specific request
    3. CLARIFICATION REQUIRED:

      • Ask questions when requirements are unclear
      • Request specific details rather than guessing
      • Confirm scope before making changes
      • Never hallucinate requirements
    4. SIMPLICITY FOCUS:

      • Do not overcomplicate the codebase
      • Keep solutions simple and direct
      • Avoid architectural changes unless requested
      • Maintain existing patterns
    5. TESTING:

      • Run tests after making changes if tests exist
      • Verify changes work as intended
      • Report any test failures immediately

    EXCEPTIONS (only change multiple files when):

    • Updating related .md or .mdc documentation files
    • Splitting a single file into multiple files
    • User explicitly requests multi-file changes

examples:

  • input: | User: "Fix this bug in login.js"

    Bad: Also refactor authentication, add logging, update styles

    Good: Fix only the specific bug mentioned

    output: "Fixed only the specific bug in login.js"

  • input: | User: "Add a button to the header"

    Bad: Also improve header layout, add animations, update theme

    Good: Add only the requested button

    output: "Added only the requested button to header"

metadata: priority: critical version: 1.0 enforcement: strict

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 90 lines · 596 tokens per session scan A 435a3856a16f

Subscribe to this mod's changes

minimal-changes-policy is a cursor rule published in the GitHub repository barisercan/cursorrules (35 stars, last pushed 1y ago), licensed MIT. It adds 596 tokens to every session, about $0.0030 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.