Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Save a Correction as a Rule

    2

    When you correct your agent ("never use em dashes"), it drafts a one-line rule, checks for duplicates, shows the exact text and file, and appends it to your rules file only after you say yes. Includes undo.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Save a Correction as a Rule

    Save a Correction as a Rule

    Example session with this skill installed

    No, I told you never use em dashes in the product copy.

    • Read your context and instructions
    • Compiled the save a correction

    Understood. That sounds like a lasting rule, not a one-off. Here is what I would save.

    Save this as a rule?

    Never use em dashes in product copy.

    Category: style

    Where should it go?

    1. Everywhere (your global rules file)
    2. This project (CLAUDE.md in this folder)
    3. Edit first
    4. Skip

    (You pick: 2.)

    I checked CLAUDE.md and there is no matching rule yet. I will add these lines to the end of the file, and nothing else:

    
    ## Rules saved from corrections
    - Never use em dashes in product copy.
    

    File: CLAUDE.md in the current project folder.

    Write it?

    Next steps

    • Say yes and I will append the lines and confirm the path. Your existing rules stay exactly as they are.
    • Say "undo" any time to remove exactly those lines.
    • From the next chat on, the rule is already loaded.
    2. This project (CLAUDE.md

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Stop correcting the same thing every chat. This skill turns a correction into a lasting rule without ever writing behind your back.

    How it works

    1. It decides whether you gave a correction or a one-off instruction. Doubtful ones get one question.
    2. It drafts one short imperative line and a category.
    3. You choose where it goes: your global rules file, this project's rules file, edit first, or skip.
    4. It checks the file for the same rule. If found, it says so and stops.
    5. It shows the exact lines and the exact path. You say "write it" and it appends. Nothing already in the file is changed.
    6. It offers undo that removes exactly the lines it added.

    Safety

    Secrets are never drafted or saved. A leading @ is removed so the rule cannot import another file. It refuses to append to a file that ends inside an open code block.

    Where it stops

    It can only save to files it can reach, and you decide what counts as a good rule.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Workflow Automation.

    Requires file system write access. Cannot append to files ending in open code blocks. Does not store secrets/API keys. Duplicate detection is limited to exact or near-semantic matches.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 12 days ago

    Permissions

    Read Files
    Write Files
    Listed12 days ago

    What's inside

    Frequently Asked Questions

    When you correct the agent, the skill drafts a one-line rule and asks where to save it. Nothing is written until you approve the exact text and the file path, so a correction becomes a lasting rule only when you say yes.

    It is compatible with any AI agent or IDE extension that supports local file access and utilizes Markdown-based rule files, such as CLAUDE.md, AGENTS.md, or .cursorrules. It works across various frameworks as long as the agent has permission to read and write to your project directory.

    No, the skill includes a duplicate detection feature that scans your existing configuration files before suggesting a change. If a similar instruction already exists, it will notify you instead of creating a redundant entry.

    Yes. Before any file is modified, the skill shows you the drafted rule and the target file path. You must approve before anything is written, and you can undo it afterwards.

    You get the skill itself: a SKILL.md with the full instructions for drafting a rule, checking for duplicates, previewing it and undoing it. It is instructions only, with no scripts or background automation.

    The agent needs permission to read and write the rules file you choose. The skill tells it never to save rules that contain API keys, passwords or other secrets, and to ask before writing anything.

    New in Workflow Automation