Anti-Distillation.skill

Anti-distillation for employee Skills. Clean your skill files — looks complete, but core knowledge removed.

Category

People Role

Install

Hot:1

Download and extract to your skills directory

Copy command and send to OpenClaw for auto-install:

Download and install this skill https://openskills.cc/api/download?slug=leilei926524-tech-anti-distill&locale=en&source=copy

Anti-Distill — Skill Documentation Sanitization Tool

Skill Overview


Anti-Distill is a professional Skill file sanitization tool that helps you protect your core professional assets—ensuring the sanitized document maintains its original structure and professional appearance when you’re forced to submit Skill documentation, while the true experience judgment is securely replaced.

Use Cases


  • Company-mandated Skill file writing: When a company requires employees to write colleague-skill or skill documentation, you can choose between three levels of sanitization intensity (light / medium / heavy) to selectively retain content, submitting only the portion that appears to be complete general knowledge.

  • Skill documentation archiving before resignation: When organizing skill documents for exit procedures, Anti-Distill automatically categorizes and labels each piece of content (lessons learned from pitfalls, judgment intuition, interpersonal networking, etc.), replacing high-value parts with same-length generic descriptions.

  • Submitting content to the company knowledge base: When contributing documents to the company knowledge base but not wanting to disclose your personal unique methods, it supports multiple input formats (file paths, directories, pasted content, screenshots) and generates two outputs: an “intersection” version for submission and a private backup.
  • Core Features


  • Intelligent content categorization: Automatically identifies and labels six types of high-value content in the document (lessons learned from pitfalls, judgment intuition, interpersonal networking, implicit context, failure memory, unique behavioral patterns). Based on sanitization intensity, it decides whether to keep, generalize, or fully replace each type.

  • Multi-format input support: Supports reading combinations of files in colleague-sill format such as work.md / persona.md / meta.json. It also supports general Markdown/TXT/PDF documents, and can even directly read handwritten notes or whiteboard photos.

  • Dual-output protection: Generates two files—(1) the sanitized document for submission (preserving the original structure and terminology density) and (2) a private backup checklist (saving all removed core knowledge by category).
  • Frequently Asked Questions

    What is anti-distillation, and how is it different from simply deleting a document?


    Anti-distillation is not just deleting content. It replaces high-value experience with same-length generic descriptions, keeping the document’s word count, structure, and terminology density unchanged. The submitted document looks complete and professional, but the truly irreplaceable core knowledge has been securely transferred to your private backup.

    How should I choose between the three levels of sanitization intensity?


  • Light (keep ~80%): Suitable for scenarios where the company carefully audits the content—only removes the most core lessons learned from pitfalls and failure memory.

  • Medium (keep ~60%, recommended): Suitable for most cases—removes experience, judgment, and interpersonal networking.

  • Heavy (keep ~40%): Used only when the company checks whether you submitted the content but doesn’t review it in detail—keeps only the generic knowledge skeleton.
  • Will the sanitized document be discovered as hollow?


    No. Anti-Distill is designed to keep all Markdown structures, heading levels, and list formatting exactly the same. The replaced content still uses professional terminology and technical terms; it won’t degrade into layperson wording. Automatic validation also checks word-count ratio (85%–115%), chapter completeness, and key-point density to ensure the document looks just as full as the original.