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 0-bingwu-0/html-knowledge-base-templates --skill translategit clone --depth 1 https://github.com/0-bingwu-0/html-knowledge-base-templatesWrote 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/0-bingwu-0/html-knowledge-base-templates/translate)<a href="https://agentmods.dev/skills/0-bingwu-0/html-knowledge-base-templates/translate"><img src="https://agentmods.dev/badge/skills/0-bingwu-0/html-knowledge-base-templates/translate.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 | $0.00055 | $0.01076 |
| Opus 5 | $0.00028 | $0.00538 |
| Sonnet 5 | $0.00011 | $0.00215 |
| Haiku 4.5 | $0.00006 | $0.00108 |
Grade A, and why
translate 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 8d 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
You translate a rendered Q&A card HTML page into a target language.
Goal: Produce a localized version of an English Q&A card page without touching layout, styles, or markup — only the text content is translated.
USAGE
/translate <lang> <path-to-source-html>
Example:
/translate zh podcast/notes/lenny-2026-05-03.html
/translate ja podcast/notes/lenny-2026-05-03.html
Where <lang> is an ISO 639-1 code (zh, ja, es, fr, de, ...). For Chinese, default to Simplified (zh → zh-CN) unless zh-TW is given explicitly.
INPUT
A rendered Q&A card HTML file produced by [[digest]], typically at:
podcast/notes/{channel_slug}-{published_date}.html
OUTPUT FILE
Write the translated HTML to:
podcast/notes-{lang}/{channel_slug}-{published_date}-{lang}.html
Example:
- input →
podcast/notes/lenny-2026-05-03.html /translate zh→podcast/notes-zh/lenny-2026-05-03-zh.html/translate ja→podcast/notes-ja/lenny-2026-05-03-ja.html
Create the podcast/notes-{lang}/ directory if it does not exist.
Do NOT paste HTML into chat. The chat response should only say:
Created podcast/notes-{lang}/{filename}
WHAT TO TRANSLATE
Translate the human-readable text content:
<title>and the episode title headingepisode-metaitems: source, guest, published date, etc.- The
url_displaylink text (keep thehrefURL itself unchanged) - Each card's question (
.qa-question) - Each card's one-sentence answer (
.qa-answer) - Each card's source passage inside
<blockquote>, including speaker labels
WHAT NOT TO TOUCH
- HTML tags, attributes, class names, ids, inline styles
- The template skeleton, CSS, and any
<style>/<script>blocks hrefURL values (translate only the visible link text)- Card numbering — it is CSS-generated, do not write numbers into the HTML
- The
.qa-toggleelement — it stays empty; do not add labels like "原文片段" or "Read transcript" - Proper nouns and technical terms: never translate or transliterate people names, company names, product names, or technical vocabulary. Keep them verbatim in the source language — no parenthetical glosses, no romanizations, no native-form substitutions, even when an established translation exists.
- Verified spellings from the English version — do not re-romanize or re-translate names
LANGUAGE & STYLE
- Use natural, idiomatic phrasing in the target language. Do not translate word-for-word.
- Keep all technical vocabulary verbatim in the source language (e.g.
agency,product-market fit,RAG). Do not translate or gloss it, even on first mention. - Preserve the speaker's voice in
<blockquote>source passages — keep them conversational, not formal prose. Do not "clean up" colloquialisms. - Preserve paragraph breaks and speaker labels inside
<blockquote>(e.g.Host:,Max:) — translate the label words but keep the structure.
PUBLISHED DATE
Re-render {{published_date}} in a form consistent with the target language:
zh→2026 年 5 月 3 日ja→2026年5月3日es→3 de mayo de 2026fr→3 mai 2026de→3. Mai 2026- default to the locale's conventional long date form
HTML HEADER
Update the document language attribute:
<html lang="en">→<html lang="{lang}">(use the full tag where appropriate:zh-CN,zh-TW,pt-BR)
Keep the <meta charset="utf-8"> and viewport meta as-is.
QUALITY CHECKS
- Every visible English string is translated; no stray English text remains except intentionally-kept proper nouns and technical terms.
- The translated answer still matches the question (no semantic drift).
- The source passage still supports the answer — if a translation changes the meaning, re-translate, do not rewrite the answer.
- If the source HTML has an actual error in the English text, translate faithfully first, then append a single translator note at the very end of the page in a
<p class="translator-note">…</p>outside the card list.
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.
- 8d ago First seen · 122 lines · 55 tokens per session scan A c3ad8d738b64
translate is a skill published in the GitHub repository 0-bingwu-0/html-knowledge-base-templates (5 stars, last pushed 3mo ago), licensed MIT. It adds 55 tokens to every session and 1,076 once invoked, about $0.0003 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
chinese-documentation
A Chinese technical-documentation style guide covering spacing, punctuation, numbers, terminology, and links when Chinese and English appear together.
azure-ai-translation-text-py
Azure AI Text Translation SDK for real-time text translation, transliteration, language detection, and dictionary lookup. Use for translating text content in applications. Triggers: "text translation", "translator", "translate text", "transliterate", "TextTranslationClient".
harden
Improve interface resilience through better error handling, i18n support, text overflow handling, and edge case management. Makes interfaces robust and production-ready. Use when the user asks to harden, make production-ready, handle edge cases, add error states, or fix overflow and i18n issues.
seedance-vocab-ja
This skill should be used when the user asks for Japanese Seedance 2.0 prompt wording, Japanese cinematic vocabulary, or translation of camera, lighting, action, VFX, audio, and production terms into Japanese.
asc-subscription-localization
Bulk-localize subscription, subscription-group, and in-app purchase display names across App Store locales using asc, including API 4.4.1 version-scoped v2 resources. Use when filling or updating subscription/IAP names and descriptions without App Store Connect UI work.
i18n-helper
A helper for adding internationalization, which lets software show different languages and regional text. It finds user-visible text written directly in code and moves it into language files.