> For the complete documentation index, see [llms.txt](https://help.sleuth.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.sleuth.io/sleuth-skills/manage/rules.md).

# Rules

A **rule** codifies a standard, convention, or constraint that should apply automatically whenever the client is working in a matching context — a file type, a directory, or a whole repository.

Where a [skill](/sleuth-skills/manage/skills.md) loads when its description matches the user's goal, a rule loads when its scope matches the current file or context. Rules are the "standing orders" of the AI client.

## When to use a rule

* Enforcing a coding convention: "Always use pytest with database reuse. Mock external HTTP calls with vcrpy."
* Requiring a specific pattern: "When adding a new Django model, also update the factories module and write a migration test."
* Reminding the client to load another asset: "Load the `frontend-design` skill whenever editing files under `frontend/`."

## Directory layout

A rule follows the same packaging conventions as a skill — a directory with a markdown file and frontmatter. The convention is `RULE.md` with a `description`, plus optional scope metadata:

```markdown
---
name: testing
description: >-
  Enforce testing best practices: load the test-writer skill, use pytest with
  database reuse, leverage fixtures and factories, and mock external calls
  with vcrpy.
globs: "**/*.py"
---

# Testing rules

- Use pytest with `--reuse-db` unless migrations changed.
- Prefer factories (`factory_boy`) over ad-hoc dict fixtures.
- Mock HTTP calls with `vcrpy`; commit cassettes alongside tests.
- Load the `test-writer` skill when writing new tests.
```

`globs` tells the client which files the rule applies to. It maps to Cursor's rule-glob format and the equivalents in other clients.

## Rules vs skills

|                 | Skill                           | Rule                                             |
| --------------- | ------------------------------- | ------------------------------------------------ |
| Loading trigger | Description matches user prompt | Scope (file type / path) matches current context |
| On by default?  | No — model decides              | Yes — whenever scope matches                     |
| Typical content | How to *do* a task              | What to *always* remember or enforce             |
| Typical size    | 1–10k tokens                    | A few hundred tokens                             |

If you find yourself writing a rule that's 20k tokens long, you probably want a skill referenced *by* a short rule instead.

## Creating a rule

Same three entry points as any other asset: the home-page assistant, the **Create → Rule** button, or `sx add ~/.cursor/rules/testing`.

## Client compatibility

Rules are supported in Claude Code, Cursor, GitHub Copilot, Gemini (CLI / VS Code / JetBrains / Android Studio), Cline, and Kiro. The rule file lands under each client's rules directory (e.g. `.cursor/rules/`, `.claude/CLAUDE.md` fragments) at install time.

## Discovered rules

Sleuth Skills can discover existing rules in connected repositories and offer to promote them to vault-managed assets. The Asset list marks these entries with `Source: <repo>` so you know where they came from. This is the fastest way to bring pre-existing `.cursor/rules/` and `.github/copilot-instructions.md` into central management.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://help.sleuth.io/sleuth-skills/manage/rules.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.