repomix-unmixer
An agent skill by daymade, from daymade/claude-code-skills. Tags: cleanup, code, conversion, documents.
What it does
Extracts files from repomix-packed repositories, restoring original directory structures from XML/Markdown/JSON formats. Activates when users need to unmix repomix files, extract packed repositories, restore file structures from repomix output, or reverse the repomix packing process.
Install
With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:
npx skills add daymade/claude-code-skills --skill repomix-unmixer
Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):
git clone --depth 1 https://github.com/daymade/claude-code-skills
cp -r claude-code-skills/repomix-unmixer ~/.claude/skills/repomix-unmixer
Safety box score
Grade C, moderate risk: 19 of 100 risk points, rated 2026-10-16. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.
This skill is a local extractor for repomix-packed repositories: it reads one packed file and writes out many files to disk. The main risk is filesystem blast radius, because the script writes extracted paths directly without sanitizing them, so a crafted repomix file can overwrite arbitrary user-writable locations; the docs also include home-directory install examples and a recursive delete snippet.
| Category | Level | Why |
|---|---|---|
| Code & shell execution | 1 (low) | It ships and runs a Python script (`scripts/unmix_repomix.py`) and the docs invoke `python3 ...` and `chmod`, but it only uses the standard library and does not evaluate downloaded code. |
| Network access | 0 (none) | No code or documentation sends data to remote hosts or downloads external content. |
| Filesystem reach | 3 (high) | The extractor writes files with `Path(output_dir) / file_path` and `open(..., 'w')` without path sanitization, so untrusted repomix paths can escape the output directory and overwrite arbitrary files. |
| Secrets & credentials | 0 (none) | The skill does not read credential stores, key files, tokens, or browser sessions. |
| Privilege & persistence | 1 (low) | It does not require sudo or services, but the installation/validation docs write into user dot-directories like `~/.claude/skills/` and `~/.claude/plugins/...`. |
| Instruction hijack surface | 0 (none) | There is no instruction to ignore safeguards, fetch follow-on instructions, or hide behavior from the user. |
| Irreversible actions | 0 (none) | It performs a local extraction task only; it does not publish, send messages, deploy, or take external actions. |
| Transparency | 0 (none) | All shipped files are readable text; there are no binaries, base64 blobs, or unpinned remote installers. |
Before installing
- Test `scripts/unmix_repomix.py` in a scratch directory with a sample repomix file containing `../` and absolute paths.
- Back up any existing `~/.claude/skills/` entries before following the install example.
- Read the troubleshooting commands before copying them, especially any `rm -rf` or `cp -r` lines.
Rated 2026-10-16 by an automated review of 5 files from the skill's repository.
Source
- Repository
- daymade/claude-code-skills (all skills from this repository)
- Path
- repomix-unmixer/SKILL.md
- Branch
- main
- Updated
- 2026-09-20
Related skills
- repomix-safe-mixer — Safely package codebases with repomix by automatically detecting and removing hardcoded credentials before packing.
- lead-magnets — When the user wants to create, plan, or optimize a lead magnet for email capture or lead generation.
- doc-to-markdown — Converts DOCX/PDF/PPTX to high-quality Markdown with automatic post-processing.
- pdf-creator — Convert markdown files to professional PDF documents with proper Chinese font support, theme system, and visual self-check.
- pdf-to-html — Converts a PDF into one self-contained, readable HTML file that preserves images, tables.
- photo-to-scanned-pdf — Two pipelines ending at a scanner-look PDF.