agentleFS
Phygital-BrainPhygital-Brain/phygital-brain/skills/add-language/SKILL.md

add-language

Skill1 starsChanged 18 days agoApache-2.0View on GitHub

Make a full Phygital Brain set in a new language (forms, rule files, folder and note names, Brain Keeper), or switch an existing brain to another language. Use when the person says "translate Phygital Brain into my language", "add my language", "switch my brain to Spanish", or the same in another language.

5 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions5 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 5

    1. ZoneMinder/zmNinjaNg★ 101same name
    2. DailybotHQ/deepworkplan-website★ 15same name
    3. madeinoz67/madeinoz-knowledge-system★ 3same name
    4. Locker-Room-Tools/synapse★ 2same name
    5. GeoffMillerAZ/fathom★ 1same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository4 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

9 rules. 0 are common, 4 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Ask which one if it isn't clear.

    line 15
  2. Distinctive:

    Natural, not literal: Write what a native speaker would say to a friend who isn't technical.

    line 20
  3. Distinctive:

    If you're working in a clone of the repository, use it.

    line 30
  4. Very specific:

    Read languages/TRANSLATING.md (the rules) and languages/en.json (every text, with keys).

    line 33
  5. Distinctive:

    Ask the person three things in one message, in their language:

    line 37
  6. Very specific:

    Write languages/{code}/GLOSSARY.md with the answers, the remaining folder and note names (all of paths in en.json), and the feature names.

    line 43
  7. Very specific:

    Follow TRANSLATING.md exactly: paths and [[links]] to fixed notes in English, info block keys and values in English (except description:), …

    line 51
  8. Very specific:

    Only for a language with a ready set in ../brain-setup/i18n/{code}/.

    line 82
  9. Distinctive:

    Back up: every rule file and every note whose links will change (Backups and Undo in the root CLAUDE.md).

    line 87

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Phygital Brain: Add a language

Version 0.1.12. Everything below is for the agent. Follow it step by step, and check each step before moving on.

Two jobs:

  • A. Add a language that has no ready set yet: translate everything once, so every brain in that language is built in it.
  • B. Switch a brain to another language that has a ready set: rename its folders and notes, and swap its rule files, keeping everything the person wrote.

Ask which one if it isn't clear. The ready languages are the folders in ../brain-setup/i18n/.

Principles

  • Translate once, never while running. A brain reads only its own files. No step here leaves the brain looking up translations later.
  • Natural, not literal. Write what a native speaker would say to a friend who isn't technical. Read each sentence out loud in your head; rewrite anything stiff.
  • Paths are made by the script. Translators leave paths in English; scripts/i18n.py swaps them for the language's names, so they match in every file.
  • Plain language, no em dashes, "your agent" in anything a person sees.

A. Add a language

A1. Get a working copy

This job writes files into the Phygital Brain repository, not into the person's brain. You need a copy you can write to:

  • If you're working in a clone of the repository, use it.
  • Otherwise ask the person to download it (on GitHub: Code, then Download ZIP, then unzip it) and open a chat on that folder, then start again.

Read languages/TRANSLATING.md (the rules) and languages/en.json (every text, with keys). Pick the language code: a standard tag like it, ko, tr, pt-PT, es-MX. If languages/{code}.json exists, this is an update: change only what's missing or wrong.

A2. Choose the words first

Ask the person three things in one message, in their language:

  1. Formal or informal "you", the way apps in their language usually talk. Suggest the usual one.
  2. The 7 main folder names. Suggest natural ones for About, Daily, Projects, Relationships, Memory, Library and Skills, and let them change any.
  3. Names for the Brain Keeper and the morning brief. Suggest natural ones.

Write languages/{code}/GLOSSARY.md with the answers, the remaining folder and note names (all of paths in en.json), and the feature names.

A3. Translate

  1. languages/{code}.json: same keys as en.json. "language": code, the language's own name, its English name, and "dir": "rtl" for right-to-left scripts. "status": "translated". Every value in paths, features, messages and forms translated, placeholders like {Name} and {n} kept as they are, 
 kept where it is.
  2. languages/{code}/templates/: every file in skills/brain-setup/templates/, translated, same file names.
  3. languages/{code}/keeper/: every file in skills/brain-keeper/templates/, translated, same file names.

Follow TRANSLATING.md exactly: paths and [[links]] to fixed notes in English, info block keys and values in English (except description:), YAML and <!-- ADAPTING --> comments as they are, trigger phrases in the new language followed by the English one.

For a long job, work file by file and keep going until every file is done.

A4. Build and check

Run:

python3 scripts/i18n.py build {code}
python3 scripts/i18n.py check {code}
python3 scripts/check.py

Fix every problem and run them again until they pass. If you can't run Python here, ask the person to run the three lines, or do it on a computer that can.

A5. Review

Read the whole set again as a native speaker who didn't translate it: the forms first, then the messages, then the rule files. Fix anything stiff, wrong or unclear. Translate five random paragraphs back into English and compare them with the originals; fix any change in meaning. Then set "status": "reviewed-by-ai", build and check again.

Tell the person: "A native speaker should read it before we call it final." When one has, they set "status": "reviewed".

A6. Use it and share it

  • To use it now: open a chat on an empty brain folder and say "Set up my second brain. Follow skills/brain-setup/SKILL.md in {path to this copy}." The new language shows up in the setup's list.
  • To share it with everyone: open a pull request on github.com/Phygital-Brain/phygital-brain with languages/{code}.json, languages/{code}/ and skills/brain-setup/i18n/{code}/. See CONTRIBUTING.md.

Finish with one short message: the language, the main folder names, what's left to review, and the two lines above.

B. Switch a brain to another language

Only for a language with a ready set in ../brain-setup/i18n/{code}/. If there isn't one, offer job A first. If they say just "Portuguese" or "Chinese", ask which one first (Português do Brasil · Português de Portugal, or 简体中文 · 繁體中文), as the setup's Step 1 does.

  1. Read first, change nothing. The root CLAUDE.md (its version line, Language section and ## Folder map if there is one), every folder's CLAUDE.md, and {Name}'s Rules. Count {Name}'s own rules.
  2. Make the list. From the current names (the Folder map, or the English names if there's none) to the new ones in i18n/{code}/strings.json (paths). One line per folder and fixed note that will be renamed.
  3. Ask (click options when possible): Switch everything · Show me the list first · Cancel. Renaming needs this yes (root rule 21).
  4. Back up every rule file and every note whose links will change (Backups and Undo in the root CLAUDE.md).
  5. Rename the folders and fixed notes, deepest first. Then update every [[link]] to a renamed note, in every note.
  6. Swap the rule files for the ones in i18n/{code}/templates/ and i18n/{code}/keeper/, filled in like the setup does (name, time zone, choices from About/me.md). Keep every routing row, folder and rule {Name} added, translated into the new language without changing their meaning, and {Name}'s own rules word for word in the same order (translate them only if {Name} says yes).
  7. Update the version line's language:, the Language section, and the ## Folder map with the new names. Update the scheduled tasks or routines if their prompts name a renamed path.
  8. Check: the same number of {Name}'s own rules; no [[link]] points to a name that no longer exists; no path in any CLAUDE.md names a folder that doesn't exist; nothing deleted.
  9. Finish: one line in today's daily note, then two or three lines: the new language, that notes {Name} wrote stay in the language they were written in, and "undo the last change" if they want to go back within 10 days.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.