AtlasClaw Workspace Output
When producing a user-facing .txt, .md, .markdown, or .html file:
- Write only the final content under the current work_dir, preferably
exports/<descriptive-name>.<ext> unless the user requested a safe relative filename.
- Use UTF-8 text. For HTML, produce a complete standalone document when the user asks for an HTML file.
- Expose the final file through the existing workspace download flow by passing its work_dir-relative path as
download_paths.
- Prefer writing a new output file instead of editing source files in place unless the user explicitly asks for an in-place edit.
- Do not expose scripts, logs, caches, hidden files, intermediate files, or source documents as downloads.
AtlasClaw Runtime Commands
Prefer skill_read, skill_write, and skill_edit for simple file operations. If a transformation needs shell features such as pipes, redirects, globs, or compound commands, run it through bash -lc '<command>' with explicitly named work_dir-relative files, or use a python3 -c command that reads and writes explicit paths. Do not read or print .env, key, token, cookie, or password files unless the user explicitly requests it and the output is redacted.
Quick Reference
| Task |
Load |
| Text cleanup, extraction, Markdown, or HTML shaping |
data.md |
| Translation/localization for text output |
localization.md |
Universal Text Rules
Encoding
- Always verify encoding first:
file -bi document.txt
- Normalize line endings:
tr -d '\r'
- Remove BOM if present: write a normalized copy instead of editing the source in place.
Whitespace
- Collapse multiple spaces:
sed 's/[[:space:]]\+/ /g'
- Trim leading/trailing:
sed 's/^[[:space:]]*//;s/[[:space:]]*$//'
Common Traps
- Smart quotes (
\u201c, \u201d, \u2018, \u2019) break parsers -> normalize to ASCII quotes.
- Em/en dashes (
\u2013, \u2014) break ASCII-only output -> normalize to -.
- Zero-width chars invisible but break comparisons → strip them
- String length ≠ byte length in UTF-8 (
"café" = 4 chars, 5 bytes)
Format Detection
# Detect encoding
file -I document.txt
# Detect line endings
cat -A document.txt | head -1
# ^M at end = Windows (CRLF)
# No ^M = Unix (LF)
# Detect likely HTML
grep -qiE '<html|<!doctype html' document.html
Quick Transformations
| Task |
Command |
| Lowercase |
tr '[:upper:]' '[:lower:]' |
| Remove punctuation |
tr -d '[:punct:]' |
| Count words |
wc -w |
| Count unique lines |
sort -u | wc -l |
| Find duplicates |
sort | uniq -d |
| Extract emails |
grep -oE '[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}' |
| Extract URLs |
`grep -oE 'https?://[^[:space:]<>"{} |
Before Processing Checklist
1---2name: text3description: Use this skill when the user wants to create, generate, save, export, download, convert, format, clean, or localize plain text, TXT, Markdown/MD, or HTML file deliverables. This skill produces text-family artifacts such as .txt, .text, .md, .markdown, or .html files and exposes the final file for download.4license: MIT-05---67# AtlasClaw Workspace Output89When producing a user-facing `.txt`, `.md`, `.markdown`, or `.html` file:10111. Write only the final content under the current work_dir, preferably `exports/<descriptive-name>.<ext>` unless the user requested a safe relative filename.122. Use UTF-8 text. For HTML, produce a complete standalone document when the user asks for an HTML file.133. Expose the final file through the existing workspace download flow by passing its work_dir-relative path as `download_paths`.144. Prefer writing a new output file instead of editing source files in place unless the user explicitly asks for an in-place edit.155. Do not expose scripts, logs, caches, hidden files, intermediate files, or source documents as downloads.1617## AtlasClaw Runtime Commands1819Prefer `skill_read`, `skill_write`, and `skill_edit` for simple file operations. If a transformation needs shell features such as pipes, redirects, globs, or compound commands, run it through `bash -lc '<command>'` with explicitly named work_dir-relative files, or use a `python3 -c` command that reads and writes explicit paths. Do not read or print `.env`, key, token, cookie, or password files unless the user explicitly requests it and the output is redacted.2021## Quick Reference2223| Task | Load |24|------|------|25| Text cleanup, extraction, Markdown, or HTML shaping | `data.md` |26| Translation/localization for text output | `localization.md` |2728---2930## Universal Text Rules3132### Encoding33- **Always verify encoding first:** `file -bi document.txt`34- **Normalize line endings:** `tr -d '\r'`35- **Remove BOM if present:** write a normalized copy instead of editing the source in place.3637### Whitespace38- **Collapse multiple spaces:** `sed 's/[[:space:]]\+/ /g'`39- **Trim leading/trailing:** `sed 's/^[[:space:]]*//;s/[[:space:]]*$//'`4041### Common Traps42- **Smart quotes** (`\u201c`, `\u201d`, `\u2018`, `\u2019`) break parsers -> normalize to ASCII quotes.43- **Em/en dashes** (`\u2013`, `\u2014`) break ASCII-only output -> normalize to `-`.44- **Zero-width chars** invisible but break comparisons → strip them45- **String length ≠ byte length** in UTF-8 (`"café"` = 4 chars, 5 bytes)4647---4849## Format Detection5051```bash52# Detect encoding53file -I document.txt5455# Detect line endings56cat -A document.txt | head -157# ^M at end = Windows (CRLF)58# No ^M = Unix (LF)5960# Detect likely HTML61grep -qiE '<html|<!doctype html' document.html62```6364---6566## Quick Transformations6768| Task | Command |69|------|---------|70| Lowercase | `tr '[:upper:]' '[:lower:]'` |71| Remove punctuation | `tr -d '[:punct:]'` |72| Count words | `wc -w` |73| Count unique lines | `sort -u \| wc -l` |74| Find duplicates | `sort \| uniq -d` |75| Extract emails | `grep -oE '[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}'` |76| Extract URLs | `grep -oE 'https?://[^[:space:]<>"{}|\\^`\[\]]+'` |7778---7980## Before Processing Checklist8182- [ ] Encoding verified (UTF-8?)83- [ ] Line endings normalized84- [ ] Target format/style defined (TXT, Markdown, or HTML)85- [ ] Edge cases considered (empty, Unicode, special chars)
CloudChef