---
title: "Steno (content.steno)"
description: "Compact representation for repeated terms."
canonical: "https://geordy.ai/formats/steno"
---
# Steno

**Compact representation for repeated terms.**

A stenography dictionary - shorthand outlines mapped to their full translations. Used to teach AI engines that every variant of a term resolves to the same canonical phrase.

Steno is the dictionary format used by the Open Steno Project (Plover) and machine stenography in general. Each entry maps a stroke or outline - the kind of compact key sequence a stenographer types - to the full word, phrase, or canonical form it expands to. Reframed as a publishing artifact, a Steno file behaves like a brand glossary in machine form: a deterministic table from every shorthand a user, crawler, or assistant might encounter to the single phrase you want them to land on.

## At a glance

- File: `content.steno`
- First released: 2010
- Created by: Open Steno Project (Plover)
- Specification: https://plover.wiki/index.php/Dictionary_format
- Read by: ChatGPT, Claude, Perplexity

## Why it matters for AI

A Steno dictionary is structurally a JSON map from outline to phrase. Borrowed from stenography and reframed for publishing, it lets you declare every acronym, internal codename, and bot user-agent you mention on the site and bind each to a single canonical expansion. LLMs and crawlers reading the file get an unambiguous lookup table for your vocabulary instead of guessing whether "GPTBot" and "GPT/PWBOT" are the same thing.

## Example

```
{
  "GEO": "Generative Engine Optimization",
  "LLMO": "Large Language Model Optimization",
  "GPT/PWBOT": "GPTBot",
  "KHRAUD/PWBOT": "ClaudeBot",
  "PER/PLEBGT/PWBOT": "PerplexityBot",
  "TKPWRAO/TKEU": "Geordy",
  "SREPBL/PWEUGT": "AI visibility"
}
```

## Benefits

- Plain JSON the same shape as a Plover dictionary - easy to author and version-control
- Every shorthand or abbreviation maps to one canonical phrase
- Acts as a brand glossary AI engines can read deterministically

## Best practices

- Keep the file as plain JSON (the Plover-native shape) so any parser can read it without a custom loader.
- Map every brand-specific shorthand, product code, acronym, and bot user-agent to its full, human-readable form.
- Treat the dictionary as the source of truth - link to it from your llms.txt so AI engines find it during ingestion.
- Version the file alongside the rest of your structured-data outputs; entries are cheap to add, expensive to drift.

## Pitfalls

- Letting the dictionary drift from the spelling used in your prose and schema - AI engines reconcile across signals and a mismatch flags noise.
- Treating it as a thesaurus instead of a glossary - one outline should map to one canonical phrase, not a list of synonyms.
- Inventing outlines that aren't real Plover strokes - the file still parses as JSON, but you lose round-trip compatibility with steno tooling.

## Use cases

- **Brand and product glossary**
- **Acronym / abbreviation expansion**
- **Bot user-agent normalization**

## Adopters

- Open Steno Project (Plover) (https://www.openstenoproject.org/plover/)
- Plover on GitHub (https://github.com/openstenoproject/plover)
---

Source: https://geordy.ai/formats/steno
This is a machine-readable markdown version of that page, generated by Geordy.