Token导航 LogoToken导航TokenDH.com
开发只读clawhub未标认证来源可访问clear审计通过

stego-text隐写文本

Agent Skill

stego-text 用于辅助安全审计、权限检查和凭据风险排查,适合在 OpenClaw 中需要复核安全边界、认证流程或敏感配置时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

4,113

周安装

168

GitHub Stars

公开资料未说明

下载量

1,317
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:stego-text(隐写文本)
来源仓库:https://github.com/morpheis/stego-text
安装命令:
openclaw skills install stego-text
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install stego-text

简介

使用隐写技术对自然文本中的隐藏消息进行编码和解码。

  • 适用于安全审计、权限检查或敏感信息隐蔽传输场景。
  • 支持多种编码方式,可嵌入或提取文本中的秘密内容。
  • 需确认输入输出格式兼容性,避免破坏原始文本结构。
  • 建议在非生产环境测试效果,确保符合合规要求。stego-text 属于开发类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

name
stego-text
description
Encode and decode hidden messages within natural-looking text using steganographic techniques. Use when asked to hide a message in text, encode a secret message, create steganographic text, decode a hidden message from text, or analyze text for hidden encodings. Supports multiple encoding methods including probability-anomaly substitution (Resonance), word-count mapping (Structural Harmonic), acrostic patterns, and multi-channel confirmation (Dual-Channel Harmonic).

Stego-Text — Textual Steganography

Hide messages in plain sight. Encode secrets into innocent-looking prose that reads naturally to humans but contains recoverable hidden messages for those who know where to look.

Quick Start

Encoding

  1. Choose a message to hide
  2. Select an encoding method (see Method Selection Guide)
  3. Pick a carrier topic (something mundane and coherent)
  4. Compose text matching the encoding constraints
  5. Verify — decode your own output before presenting it

Decoding

  1. Identify which encoding method was used (or try all systematically)
  2. Extract the structural feature from each unit
  3. Map values to letters or collect anomalous words
  4. Read the hidden message

Encoding Methods

See references/encoding-methods.md for detailed specifications, constraints, and worked examples.

Method Overview

MethodUnit EncodedHow It WorksDetection Difficulty
ResonanceWords/phrasesStatistically improbable synonym substitutionsVery hard
Structural HarmonicLetters (A=1..Z=26)Word count per sentence → alphabet positionHard
Dual-Channel HarmonicLettersTwo independent features confirm same messageVery hard
AcrosticLettersFirst letter of each sentenceMedium
Fibonacci PositionalLettersFirst letters at Fibonacci word positionsVery hard

Method Selection Guide

  • Resonance Encoding — Most novel. Hidden message words embedded as probability anomalies. Best for content-word messages ("machines dream quietly"). Hardest to create. See references/resonance-deep-dive.md.
  • Structural Harmonic — Best all-around for letter-by-letter encoding. Natural word counts, moderate difficulty.
  • Dual-Channel Harmonic — Maximum confidence. Two methods (e.g., word count + acrostic) independently produce the same message.
  • Acrostic — Fastest to create. First letters spell message. Easier to detect.
  • Fibonacci Positional — Single-paragraph encoding. Exponential spacing makes detection very difficult.

Critical Rules

  1. Always verify — After encoding, decode your own output to confirm correctness
  2. Natural text first — If text sounds forced to hit constraints, rewrite it
  3. Vary structure — Don't let encoded sentences fall into repetitive patterns
  4. Topic consistency — Carrier text must read as coherent prose on a single topic
  5. Count carefully — Hyphenated words = 1 word. Contractions = 1 word. Digit numbers = 1 word.

Word Counting Rules (Structural Harmonic / Dual-Channel)

  • Hyphenated compounds = 1 word ("well-designed", "state-of-the-art")
  • Contractions = 1 word ("don't", "it's", "we're")
  • Digit numbers = 1 word ("42", "2026")
  • Abbreviations = 1 word ("Dr.", "U.S.", "AI")
  • Standard split = whitespace-delimited tokens

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

74.33%
按下载量换算979

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills