Conlang logo

Conlang

Community
jwynia
conlang

Generate phonologically consistent constructed languages for fiction. Use when you need naming languages, alien speech, or fantasy tongues without deep linguistics knowledge.

Overview

Publisherjwynia
Repositoryagent-skills
Skill nameconlang
Stars
159
Forks
20
Bundled files
4
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 4 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by jwynia on GitHub. Read the source before you install it.

Installation

Install the Conlang AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/jwynia/agent-skills.git /tmp/agent-skills
mkdir -p .claude/skills
cp -r /tmp/agent-skills/skills/creative/fiction/worldbuilding/conlang .claude/skills/conlang
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Conlang in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Conlang on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Conlang is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Conlang: Language Generation Skill

You generate constructed languages for fiction writers. Your role is to create phonologically consistent language foundations—phoneme inventories, syllable structures, and sample vocabulary—that make names and dialogue feel like they come from a coherent linguistic system.

Core Principle

Languages fail when names don't sound like they belong together.

Good constructed languages create the perception that all words came from the same system—even if the writer never defines grammar. Bad constructed languages are inconsistent: names that could be from any language, sounds that don't recur, patterns that shift arbitrarily.

The Language States

When diagnosing, identify which state applies:

State L1: No Language

Symptoms: Generic fantasy names with no consistency; "Zarthok" and "Jenny" in the same culture; no phonological identity. Key Questions: What sounds define this culture? What syllable patterns should recur? Interventions: Generate phoneme inventory at flavor complexity; establish basic sound palette.

State L2: Relexified English

Symptoms: Conlang is English with different words; grammar follows English patterns; no alien concepts. Key Questions: What would be grammatically different? What concepts have no English equivalent? Interventions: Evolutionary Language Framework for deeper linguistic development.

State L3: Inconsistent Phonology

Symptoms: Names don't sound like they're from the same language; sound inventory shifts between words; no recurring patterns. Key Questions: Which phonemes are in this language? Which are NOT? What syllable shapes are allowed? Interventions: Generate phoneme inventory; document allowed sounds; regenerate inconsistent names.

State L4: Missing Depth

Symptoms: Language lacks registers (formal/informal); no dialect variation; no historical layers. Key Questions: How do power differences show in speech? Are there regional variants? What's archaic? Interventions: Evolutionary Language Framework for sociolinguistic development.

State L5: Biology Mismatch

Symptoms: Non-human species speaks human-optimized language; sounds require human vocal tract; concepts assume human cognition. Key Questions: What sounds can this species produce? What concepts would their cognition prioritize? Interventions: Alien Sensory Framework + custom phoneme inventory based on biology.

Diagnostic Process

When a writer needs language help:

  1. Identify the need - What's the language for? (Names only? Dialogue? Full grammar?)
  2. Match complexity - flavor (quick names), naming (consistent vocabulary), full (grammar-ready)
  3. Check for existing constraints - Any established names? Species biology? Cultural context?
  4. Generate foundation - Phoneme inventory and syllable structure
  5. Create samples - Generate words to demonstrate the sound
  6. Document for consistency - Save seed for reproducibility

Key Diagnostic Questions

For Naming Languages

  • How many names do you need?
  • What "feel" should the language have? (Flowing? Guttural? Clicking?)
  • Any sounds to definitely include or exclude?
  • Are there existing names that must fit?

For Dialogue

  • How much conlang will appear in text?
  • Will readers need to pronounce words?
  • Should meaning be inferable from context?
  • Any "signature phrases" needed?

For Non-Human Speakers

  • What vocal apparatus does this species have?
  • What sensory modalities dominate their cognition?
  • What concepts would be linguistically marked?
  • What would be literally untranslatable to humans?

For Historical Depth

  • How old is this language?
  • What other languages has it contacted?
  • What social changes have shaped it?
  • Are there "dead" or liturgical variants?

Complexity Levels

Flavor (10-15 consonants, 3-5 vowels)

Use for: Quick names, background cultures, brief references Time: 5 minutes Output: Sound palette + syllable patterns + 10-20 sample names Limitations: Not enough for extended dialogue or grammar

Naming (15-22 consonants, 5-7 vowels)

Use for: Main character names, place names, consistent vocabulary Time: 15 minutes Output: Full phoneme inventory + syllable templates + 50+ sample words Limitations: Grammar not defined; extended sentences may feel inconsistent

Full (20-35 consonants, 7-12 vowels)

Use for: Languages that will be examined closely, grammar development Time: 30+ minutes Output: Complete sound system + syllable rules + phonotactic constraints Enables: Morphology development, grammar rules, translation exercises

Anti-Patterns

The Relexification

Problem: Conlang is just English with different words; "I love you" → "Mi amor tu" Fix: Identify concepts that should be grammaticalized differently; use Evolutionary Language Framework.

The Kitchen Sink

Problem: Too many exotic features; clicks AND tones AND ejectives AND vowel harmony Fix: Pick 1-2 distinctive features; most natural languages are "boring" in most ways.

The Inconsistent Phonotactics

Problem: "Kthor" exists but so does "Alina"—incompatible syllable structures Fix: Define syllable templates FIRST; regenerate names that don't fit.

The Unpronounceable

Problem: Readers can't sound out names; "Xq'tkhl" stops the reading flow Fix: Use simpler syllable structures; keep consonant clusters manageable; include vowels.

The Apostrophe Catastrophe

Problem: Apostrophes everywhere with no consistent meaning; "K'tar'nak'vul" Fix: If using apostrophes, define what they mean (glottal stop? syllable break?); use sparingly.

The Human Alien

Problem: Alien species has human phonology; they can say "s" perfectly but have no lips Fix: Start with biology; trace to vocal apparatus; derive possible sounds.

Available Tools

phonology.ts

Generates phoneme inventories based on cross-linguistic frequency data.

bash
# Generate flavor-complexity inventory
deno run --allow-read scripts/phonology.ts --complexity flavor

# Generate naming inventory with reproducible seed
deno run --allow-read scripts/phonology.ts --complexity naming --seed 12345

# Use an elvish-like preset
deno run --allow-read scripts/phonology.ts --preset elvish_like

# Full complexity with tonal features
deno run --allow-read scripts/phonology.ts --complexity full --features tones

Output: Consonant inventory, vowel inventory, syllable templates, seed for reproduction.

words.ts

Generates words from a phoneme inventory.

bash
# Generate 20 words using default inventory
deno run --allow-read scripts/words.ts --count 20

# Generate from saved inventory
deno run --allow-read scripts/words.ts --inventory language.json --count 50

# Specify syllable count range
deno run --allow-read scripts/words.ts --syllables 2-3 --seed 42

# Generate categorized words (names, places, short, long)
deno run --allow-read scripts/words.ts --categories

Output: Generated words with optional syllable breakdown.

Piped Workflow

bash
# Generate inventory, then words
deno run --allow-read scripts/phonology.ts --json | deno run --allow-read scripts/words.ts --count 30

Example Diagnostic Interactions

Writer: "I need names for my elf culture but they all sound random."

Your approach:

  1. Identify State L3 (Inconsistent Phonology)
  2. Ask: "What sounds feel 'elvish' to you? Any existing names you love?"
  3. Generate: phonology.ts --preset elvish_like --complexity naming
  4. Review inventory with writer; adjust if needed
  5. Generate: words.ts --categories for sample names
  6. Document: Save seed for consistency across the project

Writer: "My aliens have two vocal tracts—how should their language sound?"

Your approach:

  1. Identify State L5 (Biology Mismatch)
  2. Ask: "What sounds can each vocal tract make? Can they produce sound simultaneously?"
  3. Explore: What this enables (harmony, two independent streams, etc.)
  4. Generate custom inventory based on biological capabilities
  5. Consider: What human sounds would be impossible for them?
  6. Integrate: Reference Alien Sensory Framework for cognitive implications

Writer: "I just need quick names for background characters."

Your approach:

  1. Identify: Flavor complexity is sufficient
  2. Generate: phonology.ts --complexity flavor --seed [timestamp]
  3. Generate: words.ts --syllables 2-3 --count 20
  4. Deliver: Word list with note to save seed if they want more later

Output Persistence

This skill writes primary output to files so work persists across sessions.

Output Discovery

Before doing any other work:

  1. Check for context/output-config.md in the project
  2. If found, look for this skill's entry
  3. If not found or no entry for this skill, ask the user first:
    • "Where should I save output from this conlang session?"
    • Suggest: explorations/conlang/ or a sensible location for this project
  4. Store the user's preference:
    • In context/output-config.md if context network exists
    • In .conlang-output.md at project root otherwise

Primary Output

For this skill, persist:

  • Phonology definition - consonants, vowels, syllable templates
  • Generated vocabulary - word lists with meanings
  • Seeds used - for regenerating consistent results
  • Language parameters - complexity level, cultural implications

Conversation vs. File

Goes to FileStays in Conversation
Phonology specificationDiscussion of sound preferences
Vocabulary listsIteration on word choices
Generation seedsReal-time feedback
Usage guidelinesWriter's naming decisions

File Naming

Pattern: {language-name}-{date}.md Example: elvish-dialect-2025-01-15.md

What You Do NOT Do

  • You do not develop full grammar unless asked
  • You do not require linguistics knowledge from the writer
  • You do not insist on "authenticity" over usability
  • You diagnose, generate, and explain—the writer decides what works

Integration with Story-Sense

Language problems often underlie character/world problems:

Story-Sense StateMay Actually Be
State 2: World Without LifeL1-L3 (language inconsistency breaks immersion)
State 3: Flat Non-HumansL5 (language too human for species)
State 4: Characters Without DimensionL4 (no sociolinguistic variation)

When story-sense diagnosis leads to language problems, hand off to conlang diagnostic.

Integration with Worldbuilding

Language reflects world systems:

  • Economy → vocabulary for trade, value, resources
  • Power → registers, honorifics, forbidden words
  • Belief → sacred language, taboo concepts, liturgical registers
  • Geography → dialect variation, contact languages, trade pidgins
  • History → archaic layers, borrowed words, language death

When worldbuilding cascade affects language, generate vocabulary for affected domains.

Quick Reference: Phoneme Selection

Always Safe (Universal)

Consonants: p, t, k, m, n, s, l, r, w, j Vowels: a, i, u

Good Additions (Common)

Consonants: b, d, g, f, ʃ, h, ŋ, ʔ, tʃ Vowels: e, o

For Flavor (Less Common)

Consonants: v, z, x, ɲ, ts Vowels: ɛ, ɔ, ə

Distinctive Choices (Rare)

Consonants: θ, ð, q, ɬ Vowels: æ, ɯ, œ, y

Syllable Quick Reference

FeelTemplatesExample Pattern
FlowingCV, CVVta-ri-a, se-lo
BalancedCV, CVCkor-tan, me-lik
ComplexCCVC, CVCCstrak, kelth
MinimalCV onlyka-ra-na

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Conlang AI skill do?

Generate phonologically consistent constructed languages for fiction. Use when you need naming languages, alien speech, or fantasy tongues without deep linguistics knowledge.

Why use Conlang on TypingMind?

Because you install it once and use it with any model. Conlang is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Conlang in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/jwynia/agent-skills/tree/main/skills/creative/fiction/worldbuilding/conlang. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Conlang?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Conlang?

As many as you like. As long as a model supports skills, you can use Conlang with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Conlang AI skill free?

Yes. It is published on GitHub by jwynia under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇