58
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill ahonn/dotfiles --skill repomix-explorer- SKILL.md9.9 KB
Overview
This skill uses the Repomix CLI to analyze and explore codebases, producing structured output you can search and inspect. I run repomix commands, parse the generated XML (or other formats), then surface repository structure, metrics, and pattern-based findings. Results are concise, actionable, and include next-step recommendations.
How this skill works
I determine whether the target is remote or local and run the appropriate repomix command (remote output goes to /tmp by default). I prefer XML output and use --compress for large repositories. After repomix finishes, I extract metrics, read the file tree section, run targeted grep searches for patterns, and open relevant file fragments for context before summarizing findings and recommended next steps.
When to use it
- You need a quick structural overview of a repository or directory.
- You want to find specific patterns (auth, API endpoints, models, exports).
- You need metrics: file count, token estimate, or largest files.
- Exploring a dotfiles repo (neovim, nix, tmux, git) to locate configs and plugins.
- Preparing targeted reviews or refactors and want entry points and hotspots.
Best practices
- Use --compress for repositories >100k lines to reduce token count (~70%).
- Always output remote analyses to /tmp to avoid polluting the working directory.
- Run grep pattern searches before reading large output files to narrow scope.
- Choose XML for structured analysis; use JSON if you need programmatic parsing.
- Set --output to unique paths when analyzing multiple repositories to avoid overwrites.
Example use cases
- Analyze a local dotfiles folder to list neovim config files, tmux.conf, and nix flakes.
- Explore a remote GitHub repo to count files and identify main entry points (index, init).
- Find all authentication-related code by grepping for auth/login/token patterns.
- List largest files and estimate token usage before planning an AI-based code review.
- Include only specific file types (e.g., **/*.{lua,conf,nix}) to focus the analysis.
FAQ
Use --compress for large repositories (>100k lines) to reduce token usage; skip it for tiny repos where compression overhead is unnecessary.
Where is the output file saved for remote repos?
By convention I save remote outputs to /tmp/<repo>-analysis.xml. This avoids modifying the current working directory.
Why prefer XML output?
XML preserves clear file boundaries and metadata, making it easier to grep and read structured sections like the file tree and metrics.