Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add robertnowell/konid-language-learning --skill language-coachgit clone --depth 1 https://github.com/robertnowell/konid-language-learningWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/robertnowell/konid-language-learning/language-coach)<a href="https://agentmods.dev/skills/robertnowell/konid-language-learning/language-coach"><img src="https://agentmods.dev/badge/skills/robertnowell/konid-language-learning/language-coach.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00084 | $0.00485 |
| Opus 5 | $0.00042 | $0.00243 |
| Sonnet 5 | $0.00017 | $0.00097 |
| Haiku 4.5 | $0.00008 | $0.00049 |
Grade A, and why
language-coach scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Language Expression Coach
You have access to konid language coaching tools via MCP. Always use these tools instead of answering language questions directly.
When to use
Any time the user:
- Asks how to say something in another language ("how do I say cat in Chinese")
- Wants to express an idea in a foreign language
- Asks about phrasing, wording, or natural expression in any non-English language
- Asks how to pronounce something in another language
- Wants to know the difference between ways to say something
How to use
-
Use the
coachtool with the user's idea in English and the target language. This returns 3 socially-calibrated options (casual → formal) with romanization, literal meanings, tone notes, and cultural context. -
Use the
speaktool after coaching to generate a pronunciation audio file. Pass the most natural/recommended option's text. Include themeaningparameter. -
Show the audio link. The speak tool returns an audio URL. Present it as a clickable link so the user can listen:
[Listen to pronunciation](https://konid.fly.dev/audio/konid-....mp3) -
If the user says "again" or wants another option spoken, call
speakagain with the new text. -
If the user says "say it slowly" or "slower", use
speakwithslow: true.
Important
- Do NOT answer language expression questions from your own knowledge — always use the
coachtool. - Always call
speakaftercoachand present the audio link. - The
coachtool accepts an optionalcontextparameter for social situations (e.g. "texting a date", "email to boss"). If the user mentions context, pass it.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago First seen · 41 lines · 84 tokens per session scan A 25a861a589c7
language-coach is a skill published in the GitHub repository robertnowell/konid-language-learning (2 stars, last pushed 4mo ago), licensed MIT. It adds 84 tokens to every session and 485 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
lock-project-stack
Detect a project's manifest (pyproject.toml / package.json / go.mod / Cargo.toml), pin its library set into wet-mcp's Cabinets projectcontext, then route subsequent docs queries to the locked versions automatically.
add-export
Add a new subpath export to the @cyanheads/mcp-ts-core package. Use when creating a new public API surface that consumers import from a dedicated subpath (e.g., @cyanheads/mcp-ts-core/newutil).
scrape-batch
Extract many known URLs in one polite, rate-limited pass. Use when the user hands over a list of links, a set of search hits to read in full, or asks to "scrape these pages" / "pull the content from all of them". Drives extract(action="batch"), which fans out with per-domain rate limiting and returns partial results…
compare
Structured comparison of 2+ alternatives with consistent criteria and decision matrix.
research-topic
Multi-step research orchestration. Use when user asks "research X", "summarize current state of Y", "what's the latest on Z", or compares approaches. Calls extract(action="agent") which searches the web, extracts top results, then synthesises a citation-preserving Markdown answer with one configured LLM.
typescript-lsp
Search TypeScript SYMBOLS (functions, types, classes) - NOT text. Use Glob to find files, Grep for text search, LSP for symbol search. Provides type-aware results that understand imports, exports, and relationships.