exploring-repositories
Inspect or compare an external repository using a durable local clone.
Skill by maxatwork
Chalk Labs
Need help deploying exploring-repositories?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/maxatwork-exploring-repositories
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
- Runs as
- Node.js package
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/exploring-repositories
What's inside
- Skill name
exploring-repositories- Description
- Inspect or compare an external repository using a durable local clone. Accept clone URLs, GitHub links, owner/repo names, or project names.
- License
- MIT
- Instructions
- 60 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Files (4)
- LICENSE
- SKILL.md
- openai.yaml · agents/
- explore-repo.mjs · scripts/
SKILL.md (first lines)
# Exploring Repositories
## Overview
Use a durable reference clone instead of the current thread workspace. Resolve the requested project, clone or safely refresh it under `${XDG_DATA_HOME:-$HOME/.local/share}/skillbox/explore`, then report the absolute local path so the thread can reuse it.
## Workflow
1. Resolve the target:
- Clone URL: use it directly.
- GitHub page, blob, tree, issue, or PR URL: extract `owner/repo` and use `https://github.com/owner/repo.git`.
- `owner/repo`: use GitHub HTTPS clone URL.
- Project name: use GitHub search, then ask only if the best match is not credible or the result is ambiguous for the user's request.
2. Run the bundled helper from this skill directory:
```bash
node scripts/explore-repo.mjs "<clone-url-or-github-url-or-project-name>"
```
3. If the helper reports `action: fetched-only`, use the path but do not force-reset, stash, clean, checkout, or overwrite local work unless the user explicitly asks.
4. If local Git tooling prevents cloning, continue from accessible upstream primary source at a pinned revision and disclose that no durable clone was obtained.
5. When a clone exists, copy the exact `path:` line from the helper output into follow-up commands; do not assume the default path when `XDG_DATA_HOME` may be set.
6. Continue the user's repository inspection from the reported path or pinned upstream source.
7. In the response, include:
- for a clone: `Local repo: /absolute/path`, `Clone URL: ...`, and refresh state (cloned, refreshed, or fetched-only with the reason)
- without a clone: the upstream source URL, pinned revision, and local tooling blocker
## Quick Reference
| Need | Command |
|---|---|
| GitHub URL or blob URL | `node scripts/explore-repo.mjs "https://github.com/antirez/kilo/blob/master/kilo.c"` |
| Owner/repo | `node scripts/explore-repo.mjs "vitejs/vite"` |
| Project name | `node scripts/explore-repo.mjs "difftastic"` |
| Custom clone URL | `node scripts/explore-repo.mjs "https://gitlab.com/group/project.git"` |
The helper stores repositories under a host path such as:
```text
${XDG_DATA_HOME:-$HOME/.local/share}/skillbox/explore/github.com/owner/repo
```
## Safety Rules
- Never clone reference repositories into the current thread's `work/` directory unless the user explicitly requests a throwaway clone.
- Never run destructive refresh commands such as `reset --hard`, `clean`, or forced checkout for this workflow.
- If an existing clone has local edits, untracked files, local-only commits, no upstream, or a detached HEAD, fetch remote refs only and report why the worktree was not changed.
- If the target path exists but is not a Git repo, or its `origin` points to a different repository, stop and report the conflict.
- Prefer absolute paths in follow-up commands and final responses.
## Common Mistakes
| Mistake | Fix |
|---|---|
| Using a transient path like `work/repos/<name>` | Use `${XDG_DATA_HOME:-$HOME/.local/share}/skillbox/explore/...` |
| Treating a GitHub blob URL as the clone URL | Extract only `owner/repo`, then clone the repository |
| Pulling over local notes or edits | Fetch only, preserve the worktree, and report `fetched-only` |
| Returning only the GitHub URL | Include the absolute local repo path in the response |From maxatwork/skills, MIT.
More from maxatwork/skills
- 0.589score
impeccable
maxatwork
Design or improve frontend UX and visual quality.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
model-selection
maxatwork
Choose the cheapest sufficient model and effort for coding, reasoning, taste, and throughput needs.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
penpot-canvas
maxatwork
Create, open, edit, inspect, and preview designs in a user-configured Penpot instance using Codex's built-in browser and the Penpot MCP server.
SkillAFreshness: A★ 1Claude Code +9 - 0.581score
forgejo-cli
maxatwork
Use when need to work with Forgejo repository tasks, including pull requests, issues, releases, and Actions.
SkillAFreshness: A★ 1Claude Code +9 - 0.562score
find-skills
maxatwork
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express…
SkillAFreshness: A★ 1Claude Code +9 - 0.562score
grilling
maxatwork
Grill the user relentlessly about a plan, decision, or idea.
SkillAFreshness: A★ 1Claude Code +9 - 0.552score
create-verification-skill
maxatwork
Use ONLY when the user asks to create a repository verification skill.
SkillBFreshness: B★ 1Claude Code +9 - 0.549score
fallow
maxatwork
Analyze JavaScript and TypeScript with Fallow for unused code, duplication, complexity, architecture, styling, or changed-code risk.
SkillBFreshness: B★ 1Claude Code +9 - 0.549score
diagnosing-bugs
maxatwork
Diagnose bugs or performance regressions whose cause is unclear using reproduction, targeted evidence, and hypothesis testing.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
codebase-design
maxatwork
Design or review module interfaces, ownership, and test seams using deep-module principles.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
domain-modeling
maxatwork
Change or clarify domain concepts, terminology, ownership, and lifecycles.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
ponytail-audit
maxatwork
Audit a whole codebase for unnecessary complexity when the user asks for over-engineering or bloat analysis.
SkillBFreshness: B★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 29, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/maxatwork/skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from maxatwork
- 0.589score
impeccable
maxatwork
Design or improve frontend UX and visual quality.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
model-selection
maxatwork
Choose the cheapest sufficient model and effort for coding, reasoning, taste, and throughput needs.
SkillAFreshness: A★ 1Claude Code +9 - 0.584score
penpot-canvas
maxatwork
Create, open, edit, inspect, and preview designs in a user-configured Penpot instance using Codex's built-in browser and the Penpot MCP server.
SkillAFreshness: A★ 1Claude Code +9 - 0.581score
forgejo-cli
maxatwork
Use when need to work with Forgejo repository tasks, including pull requests, issues, releases, and Actions.
SkillAFreshness: A★ 1Claude Code +9 - 0.562score
find-skills
maxatwork
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express…
SkillAFreshness: A★ 1Claude Code +9 - 0.562score
grilling
maxatwork
Grill the user relentlessly about a plan, decision, or idea.
SkillAFreshness: A★ 1Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get exploring-repositories running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
exploring-repositories skill. Agent Mart. https://agentmart.dev/skills/maxatwork-exploring-repositories (last updated 2026-10-24)