Transport diagnostics
Steganography Encoder and Decoder
This page tests how hidden transport layers can sit around an IOTA-1 (ɩ≃1) string without becoming the meaning of that string. The safest pattern is a sidecar or wrapper after conversion: keep visible ɩ≃1 canonical, scan the raw carrier, and treat hidden bytes as transport data only.
Local browser workbench
Encode, decode, and scan carriers
The workbench runs in the browser. Sidecar mode wraps unchanged visible text in inspectable markup. Zero-width mode is a fragile post-conversion test that hides a short payload after grapheme clusters and then reports every invisible carrier it used.
The public API also exposes /api/iota/steganography/scan for server-side diagnostics: original length, scalar count, grapheme estimate, zero-width carriers, hidden controls, unusual whitespace, variation selectors, bidi controls, private-use characters, cleaned visible text, and developer JSON.
Carrier output
Raw transport form
Carrier report
Run a scan to see code points, hidden-carrier warnings, and public-symbol boundary notes.
Best fit
Markup sidecar
Use a structured wrapper when the transport is HTML, XML, UAI-style records, or another envelope. The visible IOTA-1 expression remains unchanged while payload data is stored beside it and can be audited separately.
Fragile test
Zero-width postfix
Use only after conversion and only on channels that preserve the exact raw string. The scanner flags zero-width marks because they can survive visually while changing the parse surface.
Avoid
Homoglyph substitution
Do not hide data by replacing visible glyphs with confusable characters. That changes the public symbol evidence and collides with the protocol boundary.
Experimental
Semantic bucketing
Do not rely on final IOTA-1 output as a hidden-language decoder. Approximate conversion can converge different phrases into the same gist, so any source-phrase bucket must be trace-visible and reviewable.
Converter rule
Place steganography after semantic conversion.
IOTA-1 first preserves the expression, normalizes and segments it, ranks concept evidence, then emits public-symbol candidates. Hidden characters inserted before conversion can change tokenization, normalization, Unicode safety, and candidate ranking. A wrapper applied after conversion keeps the converter trace inspectable.
Use the language converter to produce a visible approximate ɩ≃1 expression and inspect evidence.
Normalize the visible text and store it as the public expression that reviewers can see.
Add sidecar metadata or a short zero-width test payload without changing the visible glyph choices.
Detect invisibles, format controls, private-use code points, and noncharacters before submitting text back into the converter.
Method comparison
Transport methods and Protocol5 fit
| Method | Fit | Reason | Guardrail |
|---|---|---|---|
| Markup sidecar | High | Visible IOTA-1 text stays unchanged. | Hash visible text and expose sidecar profile. |
| Zero-width postfix | Medium | Works only when raw strings preserve invisible marks. | Scan and strip before converter reuse. |
| Whitespace | Low | Browsers, editors, and email often trim or normalize it. | Use only in controlled files. |
| Homoglyphs | Very low | Confusables alter public symbol identity. | Reject for IOTA-1 carriers. |
| Semantic paraphrase | Experimental | Approximate ranking is not a reversible codebook. | Require source trace, not output-only decoding. |
| Image or audio wrapper | Medium | Useful when IOTA-1 is rendered as media. | Keep the source expression separately auditable. |
Do not claim