This skill is a downstream specialization of the ctf-sandbox-orchestrator skill. It covers steganography challenges in image, audio, video, document and container files: metadata, thumbnails, appended trailers, alpha and palette channels, LSBs and transcoding artifacts.
The agent confirms the real container structure first, ranks candidate channels by evidence, keeps each extracted layer as a separate file and stops when the hidden payload is reproduced. Replies are in Simplified Chinese unless English is requested.
What You Get
- A first-pass checklist for magic bytes, chunk layout, metadata, thumbnails and trailers
- A ranked list of candidate channels: alpha, palette, LSB, frame deltas, document streams and transform residue
- A rule to keep every extracted stage as its own file with its transform recorded
- A hand-off pointer to the broader crypto skill when the extracted data is ciphertext
Common Use Cases
- Looking for appended archives or trailers in an image from a CTF challenge
- Checking alpha, palette and LSB channels of a sandbox image
- Documenting the transform chain that produced a hidden payload
What Makes It Unique
It favors evidence-driven checks of structure and metadata before signal-domain work and avoids blind brute force. It separates a channel hit from artifact recovery, and it runs only after the orchestrator skill has set sandbox assumptions.
Under The Hood
The skill is instruction text only: a SKILL.md, one reference checklist and an agent config file that disables implicit invocation. It ships no scripts and does not itself fetch pages, call APIs, run commands or write files. The agent inspects media with tools the user already has, and keeps extracted layers as separate files.