Anti Gold Plating

    1

    Keep AI-assisted work within the user-approved scope. Get approval for extra improvements or real alternatives, preserve unrelated behavior, and report verification clearly.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Anti Gold Plating

    Example session with this skill installed

    Make an email report readable on mobile without changing its desktop design.

    • Read your context and instructions
    • Compiled the anti gold plating

    Scope summary: Adapt the report for mobile while preserving the desktop design, exported data, and delivery behavior. I’ll make only the mobile layout change and verify those preserved outcomes.

    I noticed an optional improvement: shorten the verbose labels across both mobile and desktop. That changes desktop copy beyond the request, so I’ll leave it out unless you approve it. If you want it, I can present the exact affected labels, benefits, risks, and effort for your decision.

    I updated the mobile layout and checked the result. The mobile report is readable at the requested viewport; the desktop appearance, exported values, and delivery behavior remain as before. No additional changes were made.

    Initial Request / Delivered: Mobile readability improved; the requested desktop design and report behavior were preserved.
    Validated Evolutions: None.
    Unvalidated Discrepancies: None observed after review.
    Verification & Limitations: I checked the changed layout and the preserved desktop view. I did not independently test other viewports or delivery systems.

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

    About this skill

    Anti Gold Plating helps an AI agent carry out requested changes without silently expanding their scope. It establishes the request and constraints, asks for explicit approval before implementing extra improvements, presents genuine alternatives for the user to choose, tracks decisions, and checks the result against the request before delivery. It is an instruction-only skill for code, configuration, documents, and systems. It does not provide sandboxing or enforce permissions, and it does not guarantee an agent will follow every instruction.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.

    This is a behavioral instruction set, not a sandbox or permission system. It cannot guarantee that an agent follows every instruction or avoids every regression. Its installation and invocation guidance targets Codex; other agents may need adaptation.

    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

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 3 days ago

    • Free with an Agensi account

    Requires

    Nothing to install.

    Permissions

    No special permissions declared or detected

    Listed3 days ago

    What's inside

    Frequently Asked Questions

    No. It instructs the agent to keep work within scope and validate changes, but it does not enforce permissions or provide a sandbox.

    No. An explicit request authorizes its stated scope. The skill calls for clarification when ambiguity changes the outcome or permitted action, and approval for additions outside scope.

    It is written and documented for Codex. Other agents that support SKILL.md can adapt it, but those integrations are not claimed as tested.

    New in Agents & Orchestration

    Free