菜单
Reference for preferred UNIX tools over LLM processing. Use when processing JSON, text, searching files, or doing batch operations.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
GDScript coding standards for The Sparkling Farce. Use when writing or reviewing code for style compliance.
Where to place game content vs platform code. Use when creating new files or moving existing code.
Run the project test suite. Use when user mentions "diagnostics", "run level 1 diagnostics", or asks to run tests.
Core development philosophy - fix platform code, not mod content. Load when making decisions about what to fix or implement.
Correct patterns for accessing game resources via ModLoader registry. Use when loading characters, items, abilities, or any mod content.
View the newest screenshot(s) from the user's Screenshots folder. Use when user says SNS, SNS2, SNS3, or asks to see a screenshot.
| name | unix-tools |
| description | Reference for preferred UNIX tools over LLM processing. Use when processing JSON, text, searching files, or doing batch operations. |
Prefer UNIX tools over LLM processing. Use the right tool for the task.
| Tool | Use For | Example |
|---|---|---|
jq | Query, filter, transform JSON | jq '.load_priority' mod.json |
| Tool | Use For | Example |
|---|---|---|
awk | Columnar data, calculations, conditional logic | awk -F: '{print $1}' file |
sed | Stream editing, find/replace, extract line ranges | sed -n '10,20p' file |
cut | Extract fields/columns by delimiter | cut -d',' -f2 file.csv |
sort | Sort lines (use before uniq) | sort -n numbers.txt |
uniq | Deduplicate adjacent lines | sort file | uniq -c |
tr | Translate/delete characters | tr '[:upper:]' '[:lower:]' |
head/tail | First/last N lines of file | head -20 file |
rev | Reverse strings | echo "hello" | rev |
| Tool | Use For | Example |
|---|---|---|
find | Locate files by name, type, date, size | find . -name "*.gd" -mtime -1 |
xargs | Pipe output as arguments to another command | find . -name "*.gd" | xargs wc -l |
| Tool | Use For | Example |
|---|---|---|
wc | Count lines (-l), words (-w), chars (-c) | wc -l *.gd |
diff | Show differences between files | diff old.gd new.gd |
comm | Compare sorted files (unique to each, common) | comm -23 file1 file2 |
| Tool | Use For | Example |
|---|---|---|
basename | Extract filename from path | basename /path/to/file.gd |
dirname | Extract directory from path | dirname /path/to/file.gd |
realpath | Resolve to absolute path | realpath ../relative/path |
| Tool | Use For | Example |
|---|---|---|
stat | File metadata (size, timestamps, permissions) | stat file.gd |
file | Detect file type by content | file mystery_file |
tree | Visualize directory structure | tree -L 2 core/ |
du | Disk usage by directory | du -sh mods/* |
| Tool | Use For | Example |
|---|---|---|
seq | Generate number sequences | seq 1 10 |
shuf | Random selection/shuffle lines | shuf -n 5 names.txt |
bc | Arbitrary precision calculator | echo "scale=2; 22/7" | bc |
date | Format dates, convert timestamps | date +%Y-%m-%d |
| Tool | Use For | Example |
|---|---|---|
md5sum/sha256sum | Compute/verify hashes | sha256sum file.zip |
base64 | Encode/decode base64 | base64 -d encoded.txt |
xxd | Hex dump for binary inspection | xxd binary_file | head |
Use UNIX tools for data processing, not LLM token consumption. These tools are faster, deterministic, and won't hallucinate results.
