← Back to Skills Marketplace
151
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install png-lsb-skills
Description
PNG图片LSB隐写分析、块信息解析与CRC校验工具，支持提取sRGB/gAMA/pHYs元数据并检测多种LSB隐藏模式；适合CTF比赛场景，当用户需要分析PNG图片结构、验证数据完整性、检测隐写信息或提取图像元数据时使用
Usage Guidance
该技能在功能声明、说明文档与实际脚本实现上是一致的：它在本地读取指定PNG文件，解析chunk并做CRC校验，使用Pillow提取像素以分析前8行的LSB位，且没有网络调用或读取其它系统凭据。尽管如此，安装与运行任何第三方脚本前仍建议：1) 在受限/沙箱环境中运行以防意外（例如临时容器或虚拟环境）；2) 通过受信任渠道安装依赖（pip 从官方源）并确认 Pillow 版本；3) 如果需要更深入的隐写检测（超过前8行或需要自动化大量文件），审阅并必要时修改脚本参数/代码；4) 如需完全保证安全，可让有经验的人审查代码以确认无隐蔽行为（当前脚本未发现网络/外发逻辑）。
Capability Analysis
Type: OpenClaw Skill
Name: png-lsb-skills
Version: 1.0.1
The skill bundle is a legitimate PNG analysis tool designed for CTF (Capture The Flag) scenarios, focusing on chunk parsing, CRC validation, and LSB steganography detection. The core logic in `scripts/png_analyzer.py` uses standard libraries (Pillow, binascii, struct) to parse image data and extract bit patterns without any network activity, unauthorized file access, or command execution. The instructions in `SKILL.md` are strictly aligned with the stated purpose of image analysis.
Capability Assessment
Purpose & Capability
技能名与描述（PNG签名、chunk解析、CRC校验、元数据提取、LSB隐写分析）与包含的脚本（scripts/png_analyzer.py）的实际功能一致。所需依赖仅为 Pillow（在SKILL.md中声明），未请求与功能无关的凭据或系统访问。
Instruction Scope
SKILL.md 明确指示通过 `python scripts/png_analyzer.py --png <图片>` 运行脚本。脚本仅读取指定PNG文件、可选地写入指定输出JSON，并使用Pillow打开图像、解析chunk并提取前8行的LSB位。没有指示读取其它系统文件、环境变量或向外发送数据，行为与文档中描述的分析范围一致。
Install Mechanism
没有安装规范（instruction-only 安装），仅包含 Python 脚本和说明。唯一依赖是 Pillow（公开的Python包），没有下载未知二进制或从不可信URL提取代码。风险低且与功能相称。
Credentials
技能不要求任何环境变量、凭据或配置路径，脚本也没有访问环境变量或其他凭据。所请求的权限与其处理本地图片文件并可写入输出JSON的需求相称。
Persistence & Privilege
flags 表示 always:false、user-invocable:true，且 disable-model-invocation 为默认（允许模型在合适情形下调用）。技能不会修改其他技能或系统配置，也不要求持续驻留或提升权限。
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install png-lsb-skills - After installation, invoke the skill by name or use
/png-lsb-skills - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- 修正了命令和路径描述，从 /workspace/projects/png-analyzer/scripts/ 统一改为 scripts/，便于通用本地环境调用
- 未涉及功能和依赖性变更，分析流程、技能用途和示例用法保持一致
- 文档描述更简化、明确，减少冗余路径，提升易读性
v1.0.0
Initial release of png-lsb-skills.
- Provides detailed analysis of PNG image structure, including signature validation and chunk parsing.
- Performs CRC integrity checks on all PNG chunks.
- Extracts sRGB, gAMA, and pHYs metadata from PNG files.
- Detects and decodes potential LSB steganography using 15 different modes.
- Outputs analysis results as JSON or prints directly to console.
- Designed for scenarios requiring structure analysis, data integrity verification, steganography detection, or metadata extraction from PNG images.
Metadata
Frequently Asked Questions
What is png-lsb-skills?
PNG图片LSB隐写分析、块信息解析与CRC校验工具，支持提取sRGB/gAMA/pHYs元数据并检测多种LSB隐藏模式；适合CTF比赛场景，当用户需要分析PNG图片结构、验证数据完整性、检测隐写信息或提取图像元数据时使用. It is an AI Agent Skill for Claude Code / OpenClaw, with 151 downloads so far.
How do I install png-lsb-skills?
Run "/install png-lsb-skills" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is png-lsb-skills free?
Yes, png-lsb-skills is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does png-lsb-skills support?
png-lsb-skills is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created png-lsb-skills?
It is built and maintained by 末心 (@moxin1044); the current version is v1.0.1.
More Skills