agentleFS
motola10motola10/claude-code-base/.claude/skills-disabled/claude-guideline-creation/SKILL.md

claude-guideline-creation

Skill1 starsChanged 5 months agoMITView on GitHub

Creates new guideline documents or modifies existing ones, placing content into the right section without breaking formatting. Use this skill whenever the user wants to: add information to an existing guideline or cheatsheet, create a new guideline or reference doc from scratch, update a section of a standards doc, log a new rule or convention, or expand a docs/ or guidance/ file with new context. Trigger on phrases like "add this to the guideline", "update the cheatsheet", "create a new guideline for X", "log this in the docs", "add a section for Y", or any request to write or modify a structured reference document.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If ambiguous, ask: *"Should I add this to an existing guideline, or create a new one?"*

    line 32
  2. Distinctive:

    Do not ask for information the user already provided.

    line 47
  3. Distinctive:

    Read the full file

    line 53
  4. Distinctive:

    If no section fits, create a new one following the existing heading style (numbered H2/H3 for Engineering Guideline.

    line 55
  5. Distinctive:

    Write the updated file back to disk

    line 61

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

What this skill does

Manages guideline documents in two modes:

  • Modify: reads an existing file, finds the right section, inserts content cleanly
  • Create: generates a new document using a consistent, predefined format

Step 1 — Detect intent

Determine which mode applies:

SignalMode
User references an existing file path or filenameModify
User says "add to", "update", "log this in"Modify
User says "create", "new guideline", "new cheatsheet"Create
No file referenced, topic is newCreate

If ambiguous, ask: "Should I add this to an existing guideline, or create a new one?"


Step 2 — Gather missing info (only what's needed)

For Modify — if not already clear:

  • Which file? (ask for path or name if not given)
  • What content to add?

For Create — if not already clear:

  • What is this guideline for? (topic / title)
  • Which format: Engineering Guideline (numbered sections, detailed rules) or Cheatsheet (tables-first, quick reference)?
  • Any initial content to include?

Do not ask for information the user already provided. One question at a time if clarification is needed.


Step 3a — Modify an existing guideline

  1. Read the full file
  2. Identify the best-fit section for the new content based on topic proximity
  3. If no section fits, create a new one following the existing heading style (numbered H2/H3 for Engineering Guideline; plain H2 for Cheatsheet)
  4. Insert the content using the same formatting patterns already in the file:
    • New command/flag → add a table row
    • New rule or convention → add a bullet point under the relevant sub-section, using * **Bold Key:** format
    • New concept with sub-points → add an H3 sub-section with bullets
    • New major topic → add an H2 section with a horizontal rule before it
  5. Write the updated file back to disk

Formatting rules to preserve:

  • Horizontal rules (---) between major H2 sections
  • Table column alignment (:--- for left-align)
  • Blockquotes (>) for rules/tips
  • Code blocks (```) for commands and JSON
  • Bullet rules use * **Bold Key:** description — never plain prose paragraphs

Step 3b — Create a new guideline

Choose the template based on format selection.

Engineering Guideline template

# [Title]

[One-sentence description of what this document covers.]

---

## 1. [Section Name]

[Brief framing, 1 line max.]

### 1.1 [Sub-Section]

* **Rule or concept:** Description.
* **Another rule:** Description.

```code example if relevant```

---

## 2. [Section Name]

### 2.1 [Sub-Section]

| Column A | Column B |
| :--- | :--- |
| Item | Description |

---

## Key Resources

| Resource | URL |
| :--- | :--- |
| [Name] | [URL] |

Cheatsheet template

# [Title] Cheatsheet

---

## [Section Name]

| Command / Setting | Description |
| :--- | :--- |
| `command` | What it does |

> **Rule:** Key rule or tip.

---

## [Section Name]

| Command | Description |
| :--- | :--- |

---

## Key Resources

| Resource | URL |
| :--- | :--- |
| [Name] | [URL] |

Content rules for new documents

  • No filler openers: never write "This section covers...", "It is important to note that...", "In this document..."
  • Prefer tables over prose for any list of commands, flags, settings, or options
  • Prefer bullets over paragraphs for rules and conventions — always use * **Bold Key:** description format
  • Code blocks for all commands, JSON, file paths, and code snippets
  • Blockquotes for rules or tips that should stand out
  • Each entry should be one clear statement — no justification unless non-obvious

Step 4 — Save the file

  • Modify: write back to the same path
  • Create: save to the most appropriate location:
    1. A Claude Code Guidance/, docs/, or guidelines/ folder if one exists nearby
    2. Otherwise, the current working directory
    • File name: lowercase with hyphens, e.g. git-workflow-guideline.md or docker-cheatsheet.md

After saving, confirm the file path to the user in one line.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.