FlameWolf

How it works: Your message is compressed with DEFLATE, encrypted with AES-GCM using a random initialization vector (IV), then the IV and ciphertext are combined and encoded in Base-95. Decryption requires the exact key that was used to encrypt.

Features

• Modern, secure AES-256-GCM encryption.
• Auto-generated key, displayed as text and a QR code for quick sharing.
• DEFLATE compression to shrink messages before encrypting.
• File upload/download, clipboard support, byte counters and toast feedback.

Limitations

• The key is ephemeral and changes on every reload — you must save and securely transmit it, or the message is unrecoverable.
• Only text files can be uploaded.
• Trailing spaces in the output are fragile and can be lost in transit (the tool warns you).

How it works: A 256-bit AES-GCM key is derived from your passphrase using PBKDF2 (SHA-256, 250,000 iterations) and a random salt. The message is compressed with DEFLATE, encrypted with a random IV, and the salt, IV and ciphertext are encoded together in Base-95. Decrypting requires the same passphrase.

Features

• Passphrase-derived keys — nothing extra to transmit or store.
• AES-256-GCM with a fresh salt and IV for every message.
• DEFLATE compression, file I/O, clipboard support and byte/ratio statistics.
• Keyboard shortcuts (Ctrl/Cmd + Enter) and full keyboard tab navigation.

Limitations

• Security depends entirely on passphrase strength; there is no strength meter.
• Only text files can be uploaded.
• Trailing spaces in the output are fragile (the tool warns you), and forgetting the passphrase makes the message unrecoverable.

How it works: To encode, the text is compressed with DEFLATE and then the bytes are mapped to a string using a 95-character set of printable ASCII characters. Decoding reverses the process — decode from Base-95, then decompress.

Features

• Tabbed Encode / Decode interface.
• Type, paste, upload a text file, or insert a sample as input.
• Copy the output to the clipboard or download it as a .txt file.
• Live byte counts and size-change percentage to gauge compression.
• Warns when the output has trailing spaces; runs fully client-side.

Limitations

• The Base-95 character set may not be safe on every transmission channel.
• Trailing spaces are fragile, and uploads must be text files.

How it works: Malayalam characters (U+0D00–U+0D7F) and ASCII characters are mapped to a compact single-byte representation, compressed with DEFLATE, then encoded in Base-95. Decoding reverses each step.

Features

• Range-compression optimized specifically for Malayalam text.
• Normalizes some symbols (smart quotes, dashes) to their ASCII equivalents.
• Inherits the full Base-95 UI: file I/O, clipboard support and statistics.

Limitations

• Only ASCII and Malayalam ranges are supported; other characters become [UNK].
• Base-95 encoding is not encryption.

How it works: The input is normalized to standard character forms, then the selected scheme's substitution map is applied, matching multi-character conjuncts before single characters.

Features

• Two schemes (standard and simplified), selectable with radio buttons and remembered via localStorage.
• Clipboard paste/copy, sample-text insertion and toast feedback.

Limitations

• Not cryptographically secure — it is a simple substitution scheme.
• It can't distinguish dental ന from alveolar ഩ, so output may not match pronunciation. (This doesn't affect the involutory nature of the transform: applying it twice returns the original input.)
• When ര്യ is written with the Dot Reph (ൎ) character, the resulting edge case is not handled.

How it works: Malayalam character variants are normalized, then a fixed substitution map is applied to each character. Most mappings are reversible pairs, with a few one-way transforms.

Features

• Minimalist input/output layout with a Convert button.
• Clipboard paste/copy, sample-text insertion and toast feedback.

Limitations

• Malayalam only; whitespace and punctuation pass through unchanged.
• A few one-way transforms cannot be reversed automatically.

How it works: After normalizing character variants, a bidirectional map swaps vowels with paired vowels, consonants with paired consonants, and numerals with their counterparts.

Features

• Minimalist input/output layout with a Convert button.
• Clipboard paste/copy, sample-text insertion and toast feedback.

Limitations

• Malayalam only.
• One one-way transform (ൗ→ൈ) cannot be reversed automatically.

How it works: The input is normalized and tokenized (matching conjuncts first), then each token is mapped to its equivalent in the other script. Most mappings work both ways, though a few are one-way.

Features

• Input and read-only output with a Convert button.
• Clipboard paste/copy, sample-text insertion and toast feedback.

Limitations

• A few transforms (ൟ, ൗ, ൎ) are one-way and cannot be converted back from Brahmi.

How it works: The tool uses Intl.Segmenter to split text into grapheme clusters and reverses the clusters, so combining marks and ligatures stay attached to their base characters.

Features

• Live output that updates as you type.
• Clipboard paste/copy, sample-text insertion and toast feedback.

Limitations

• No word- or character-level reversal options.
• Does not reorder bidirectional (RTL/LTR) text semantically.

How it works: Using Intl.Segmenter to find grapheme boundaries, the tool swaps the first letter-grapheme between each pair of consecutive words while preserving sentence boundaries.

Features

• Live output that updates as you type.
• Works across scripts (Latin, Malayalam and more).
• Clipboard paste/copy, sample-text insertion and toast feedback.

Limitations

• Only swaps within consecutive pairs, so an odd trailing word is left unchanged.
• Words containing no letters are skipped; requires Intl.Segmenter support.

How it works: Letters A-Z are shifted 13 places and digits 0-9 are shifted 5 places. Because both shifts are self-inverse, applying ROT-18 a second time restores the original text.

Features

• Live transformation that updates instantly as you type.
• Clipboard paste/copy, sample-text insertion and toast feedback.

Limitations

• Weak obfuscation, not cryptographic security.
• Only letters and digits are transformed; everything else passes through.