refactor-safety

A set of rules for changing existing code while keeping its behavior unchanged. It requires understanding why code exists, separating cleanup from new features, and asking before deleting code or dependencies.

In plain words
What is it for?
Use it when CodeCortexLoop applies fixes or refactors, especially when simplifying code, changing performance, removing files or dependencies, or modifying error handling and configuration.
Why use it?
It helps prevent refactoring—the restructuring of code without intended behavior changes—from accidentally removing compatibility, edge-case handling, error behavior, or important side effects.

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/whitequeen306/code-cortex-loop/refactor-safety
Clone the repo
git clone --depth 1 https://github.com/whitequeen306/code-cortex-loop
Per session 37 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 858 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00037 $0.00858
Opus 5 $0.00018 $0.00429
Sonnet 5 $0.00007 $0.00172
Haiku 4.5 $0.00004 $0.00086

Measured yesterday against content hash b9e82cf5efd3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

refactor-safety 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 yesterday.

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.

rules/refactor-safety.mdc · 108 lines

How it starts

The opening of the file, as written. The whole thing — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Refactor Safety

Apply these constraints whenever /cortexloop modifies code (Direct mode or post-Report confirmation).

Core Principles

1. Preserve Behavior Exactly

Simplification and performance changes must NOT alter:

  • Inputs, outputs, or return values
  • Error behavior and error messages (unless fixing a security leak)
  • Side effects and their ordering
  • Edge case handling

Test: If existing tests need modification to pass after a "refactor", you likely changed behavior — revert and reconsider.

2. Chesterton's Fence

Before removing or simplifying anything, understand why it exists:

- What calls this code? What does it call?
- What edge cases does it handle?
- Why was it written this way? (git blame, comments, PR history)
- Is this dead, or intentionally kept for backwards compatibility?

If you cannot answer these, do not remove or simplify — flag as UNCERTAIN in the report.

3. Ask Before Deleting

Never silently delete:

  • Functions, classes, files, or exports
  • Dependencies (npm/pip/cargo packages)
  • Feature flags or config keys
  • Error handling branches
  • Comments that explain "why" (only remove "what" comments)

Use the ask-before-delete template from dead-code-and-deps skill.

4. Separate Refactors from Features

In Direct mode:

  • Refactor commits/changes: simplify, rename, extract, dead code removal
  • Fix commits/changes: bug fixes, security patches, test additions
  • Do not mix unrelated refactors with bug fixes in the same edit group
  • If a fix requires a refactor, do the fix first, verify tests, then refactor

5. Incremental Changes

  • One simplification or optimization at a time
  • Run tests after each change
  • Revert immediately on failure — do not "fix forward" blindly
  • Keep diffs reviewable — avoid 500+ line drive-by refactors

Forbidden Actions

  • Empty catch blocks (even if "cleaner")
  • Removing error handling to simplify code
  • Disabling tests to make the pipeline pass
  • @ts-ignore / # noqa / allow() to silence real issues
  • Disabling security controls as a "fix"
  • Force-pushing or rewriting git history
  • Modifying lock files without running the package manager
  • Major dependency version bumps without explicit user approval

Read the full file on GitHub · 108 lines

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. yesterday First seen · 108 lines · 37 tokens per session scan A b9e82cf5efd3

Subscribe to this mod's changes

refactor-safety is a cursor rule published in the GitHub repository whitequeen306/code-cortex-loop (15 stars, last pushed 1mo ago), licensed MIT. It adds 37 tokens to every session and 858 once invoked, about $0.0002 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.