A command-line tool for hiding and extracting UTF-8 text/binary data into images using RGB LSB steganography with capacity and integrity checks.

• Encode text/binary data into images: Uses least significant bits of RGB channels.
• Decode embedded data: Recover hidden messages from compatible images.
• File or inline input: Provide data directly or from a file.
• Optional encryption: Store a key/nonce pair in files to encrypt data before embedding.
• Safe error handling: Clear error messages and non-zero exit codes on failure.

The tool is driven by subcommands:

• encode: Embed data into a file.
• decode: Extract data from a file.
• cap: Calculate the maximum possible payload size for an image.

Run the following to see the built-in help:

cloakimg --help

• Hide a short note:

cloakimg encode data/tp0n3p08.png -t "Meet at 19:30." # output is "a.png"

• Embed the source code into a image:

cloakimg encode data/tp0n3p08.png --file src/stego/encode.rs"

• Recover a message to the terminal:

cloakimg decode data/tp0n3p08_secret.png

• Recover a message to a file:

cloakimg decode data/with_payload.png --output recovered.txt

Enable encryption for your payload prior to embedding by using a key file. The key file may consist of either 32 raw bytes or a 32-byte ASCII hex string.

A fresh nonce is generated for each encryption and automatically embedded in the payload. The format is: [12-byte nonce][N-byte ciphertext][16-byte tag]. During decryption, the nonce is extracted automatically.

# Encrypt before embedding
cloakimg encode data/tp0n3p08.png -o data/tp0n3p08_secret.png \
  --key-file secrets/image.key \
  -t "Meet at 19:30."

# Provide the same key to decrypt (nonce is extracted automatically)
cloakimg decode data/tp0n3p08_secret.png \
  --key-file secrets/image.key

Mismatched keys will prevent successful decryption. Each encryption automatically uses a unique nonce.

• Supported image formats: PNG, BMP, TIFF, PPM.
• Available capacity depends on image dimensions and encoding details; very long messages may not fit into small images.