diff --git a/README.md b/README.md index 2485c18..2816c66 100644 --- a/README.md +++ b/README.md @@ -93,9 +93,9 @@ folder, so the set stays small as coverage grows. | Skill | Category | Install name | Works on | |---|---|---|---| -| Bookmarks | browser | `organizing-bookmarks` | Ten browsers. Dedupes across tracking parameters, `www`, and trailing slashes. Prunes dead links. Flattens folder sprawl. | -| Folders | files | `organizing-folders` | Downloads, Desktop, Documents, or any folder you name. Finds byte-identical duplicates, archives stale files, groups by kind or project. | -| Notes | notes | `organizing-notes` | An Obsidian vault or any folder of markdown. Finds orphans, stubs, and near-duplicates. Consolidates tag sprawl without breaking `[[wikilinks]]`. | +| Bookmarks | browser | `organizing-bookmarks` | Every major browser. The same page saved four times under slightly different links, bookmarks that no longer go anywhere, folders holding one thing. | +| Folders | files | `organizing-folders` | Downloads, your Desktop, or any folder you name. The same file saved twice, things you have not opened in months, and the huge items you forgot were there. | +| Notes | notes | `organizing-notes` | An Obsidian vault or any folder of notes. Notes you started and never finished, two versions of the same list, tags that all mean the same thing. | ### Which one should I use? diff --git a/scripts/sync-skills.py b/scripts/sync-skills.py index b6eb9a8..2e8bf3b 100755 --- a/scripts/sync-skills.py +++ b/scripts/sync-skills.py @@ -73,7 +73,7 @@ def published(check_only): site = open(site_path, encoding="utf-8").read() new_site = re.sub(r'(class="stat">)\d+(<)', r"\g<1>%d\g<2>" % total, site) chips = "".join('%s' % a for a in chromium) - new_site = re.sub(r'(
Dedupes across tracking parameters, www, and trailing slashes. Prunes dead links. Flattens folder sprawl.
The same page saved four times under slightly different links. Bookmarks that no longer go anywhere. Folders holding one thing. It keeps one of each and sorts the rest into topics.
Downloads, Desktop, Documents, or any folder you name. Finds byte-identical duplicates, archives what has gone stale, and groups what is left by kind or by project. Names every multi-gigabyte item individually and leaves this week's work alone.
+Point it at Downloads, your Desktop, or any folder you like. It finds the same file saved twice, sets aside what you have not opened in months, and tells you about the huge things you forgot were in there. Anything you touched this week is left alone.
An Obsidian vault or any folder of markdown. Finds orphans, stubs, and near-duplicates. Consolidates tag sprawl without breaking your [[wikilinks]], and never merges two notes without asking about that specific pair.
Notes you started and never finished. Two versions of the same list. Half a dozen tags that all mean the same thing. It gathers them together without breaking the links between them, and never merges two notes without asking you first.
Skills target a category, not an app. Paths live in a shared reference file, so a skill covers every app that shares a storage format.
+One skill covers a whole kind of app rather than a single one. Show it where another browser keeps its bookmarks and it works there too, with no new skill to write.
Adding a browser takes three lines and one command.
-A row in the reference table, the data path, the process name, then one sync command. No new skill and no new code, and every browser skill covers it immediately. Every browser listed above was added that way.
The same holds across operating systems. Chrome keeps identical bookmark JSON on Windows, macOS, and Linux, so only the path changes. macOS is the strictest of the three and the one that needs a permission grant.
+Where it keeps its bookmarks, what it is called when running, and one command to tie it together. No new skill and no code at all, and every browser skill covers it straight away. Every browser above was added that way.
+The same is true across Windows, Mac, and Linux. A browser keeps your bookmarks in exactly the same way on each, so only the location changes. Mac is the strictest of the three and the only one that asks for permission first.
No build step, no API, no TypeScript. Write a markdown file and open a pull request.
+No build step and no code. Describe what tidy looks like for your app in a text file, and open a pull request.
Add a row to the reference table with the path and permissions. Every existing skill in that category covers your app immediately.
+Add a row saying where it keeps its data. Every skill of that kind then covers your app straight away.
One file at skills/<category>/<name>/SKILL.md. Point at the shared rules, then write what "organized" means for your app.
One text file. Point at the shared safety rules, then describe what tidy means for your app in plain words.
Add a parser to scripts/. Python 3 standard library only, no dependencies, and it refuses to write without a backup.
This one needs a little code to read its files. Nothing to install, and it refuses to write anything without a backup saved first.
There is no CI gate. These skills move real files, so a green check would be false comfort. Someone reads the markdown instead.
+Nothing here is approved by a robot. These skills move real files, so a green tick would be false comfort. A person reads every one.
Numbers, not adjectives. Asking a question is never consent to rewrite something.
+Numbers, not adjectives. Asking a question is never permission to change anything.
The scripts will not write without a real backup, or into a live browser that would clobber the edit on quit.
+It will not touch anything without a backup already saved, or work on a browser that is still open and would overwrite the changes when it closes.
No symlink escapes. Keys, credentials, and library data are never scanned or listed.
+It cannot slip out through a shortcut. Keys, passwords, and system data are never even looked at.