| name | anti-distill |
| description | Use when sanitizing skill files, runbooks, internal guides, prompt libraries, or tacit workflow documents before sharing them with an employer, client, or team knowledge base. 当你需要在对外共享前清洗技能文档、操作手册、内部指南、提示词库或隐性流程知识时使用。 |
Anti-Distill
反蒸馏
Overview
概述
This is the Codex/ECC discovery entrypoint for the canonical implementation in:
这是 Codex/ECC 的发现入口,真实实现位于:
anti-distill-skills/anti-distill/
Canonical Files
正式文件
Use the canonical files from this repo root:
请读取仓库根目录中的正式文件:
Preferred Execution
推荐执行方式
Preferred execution path:
推荐执行方式:
python .\anti-distill\scripts\anti_distill.py --input <path> --output-dir <dir> --level medium --interactive
Commands Quick Ref
新子命令速查区
Use these subcommands with anti-distill/scripts/anti_distill.py:
使用 anti-distill/scripts/anti_distill.py 时可用以下子命令:
analyze - inspect leverage categories, doc type, and risk
analyze - 分析杠杆类别、文档类型和风险等级
python .\anti-distill\scripts\anti_distill.py analyze --input <path>
plan - generate a defense strategy before rewriting
plan - 在改写前生成防护策略
python .\anti-distill\scripts\anti_distill.py plan --input <path> --mode sanitize
sanitize - generate cleaned output + private backup + report
sanitize - 生成清洗版 + 私有备份 + 报告
python .\anti-distill\scripts\anti_distill.py sanitize --input <path> --output-dir <dir> --level medium --interactive
generate-decoy - generate outward-facing decoy package
generate-decoy - 生成对外展示诱饵包
python .\anti-distill\scripts\anti_distill.py generate-decoy --input <path> --output-dir <dir>
- Legacy compatibility - no subcommand defaults to
sanitize
- 兼容旧用法 - 不写子命令时默认走
sanitize
python .\anti-distill\scripts\anti_distill.py --input <path> --output-dir <dir> --level light
Discovery
自动发现
If this project is the current Codex workspace root, the .agents/skills/anti-distill folder makes the skill auto-discoverable.
如果当前 Codex 工作区根目录就是这个仓库,那么 .agents/skills/anti-distill 会让技能自动被发现。