install-tooling
Use when installing new skills or agents from curated lists or GitHub repositories, when the user says "install", "add skill", "add agent", or wants to set up…
Skill by bzellman
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/bzellman-install-tooling
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
Use cases: agent-building
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
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/install-tooling
What's inside
- Skill name
install-tooling- Description
- Use when installing new skills or agents from curated lists or GitHub repositories, when the user says "install", "add skill", "add agent", or wants to set up new tooling in their Claude Code environment
- License
- MIT
- Instructions
- 249 lines in SKILL.md · view on GitHub
Before installing
- Instructions call the network
Agent Skills spec issues
- files: referenced but missing: references/project-overrides.md
Files (4)
- SKILL.md
- openai.yaml · agents/
- audit-checklist.md · references/
- curated-sources.md · references/
SKILL.md (first lines)
# Install Tooling
Install skills and agents with project-aware curation. Handles one or many items per invocation.
**Sources:** Curated lists (configurable in `references/curated-sources.md`) and ad-hoc GitHub URLs.
**Pipeline:** Intake → Audit → Install → Patch → Document
**Supports:** Both skills (directory with `SKILL.md`) and agents (single `.md` file with frontmatter). Both go through the same pipeline.
## Phase 1: Intake & Classify
Parse user input into a list of items.
**From curated list name:**
1. Read `references/curated-sources.md` for configured repos.
2. For each source, fetch the directory listing via GitHub API:
```bash
curl -s "https://api.github.com/repos/{owner}/{repo}/contents/{path}?ref=main" \
-H "Accept: application/vnd.github.v3+json"
```
3. Match the requested name against available items.
4. If not found in any source: report "not found in curated sources — provide a GitHub URL."
**From GitHub URL:**
1. Parse URL to extract `owner`, `repo`, `ref`, and `path`.
- URL format: `https://github.com/{owner}/{repo}/tree/{ref}/{path}`
- Short format: `owner/repo` + `--path path/to/skill`
2. Fetch and check for `SKILL.md` (→ skill) or single `.md` with `name:`/`description:` frontmatter and `tools:` field (→ agent).
**Listing mode:** When user asks "what's available?" without specifying items:
1. Fetch all curated sources.
2. Check each item against installed skills/agents — annotate `(installed)`.
3. Present grouped by source:
```
Skills from openai/skills (.curated):
1. skill-a
2. skill-b (installed)
3. skill-c
Skills from openai/skills (.experimental):
1. exp-skill-a
2. exp-skill-b (installed)
Which ones would you like to install?
```
**Output per item:** `{name, type: skill|agent, source: curated|github, repo, path, ref}`
## Phase 2: Audit
For each item, run three checks. Read `references/audit-checklist.md` for detailed criteria.
### 2a: Redundancy
Scan installed skills/agents in all four locations:
- `~/.agentConfig/skills/`
- `~/.agentConfig/agents/`
- `.claude/skills./` (if exists)
- `.claude/agents./` (if exists)
Read each item's YAML frontmatter and first 50 lines of body. Classify overlap as: exact duplicate, superset, partial overlap, or complementary. See `references/audit-checklist.md` for classification rules and examples.
### 2b: Value Alignment
Read `CLAUDE.md` to extract the project's tech stack. Compare against the item's description and domain. If the item targets technologies not in the project, produce a soft advisory. Never block — just inform.
### 2c: Conflict Detection
Check if the item's guidance contradicts:
- `CLAUDE.md` conventions and gotchas
- `agent-os/standards/` files matching the item's domain
See `references/audit-checklist.md` for common conflict patterns.
### Audit Report
Present one report per item:
```
Audit Report for: [name] ([skill|agent])
─────────────────────────────────────────
Redundancy: [NONE | OVERLAP with X (type) — recommendation]
Value: [ALIGNED | ADVISORY: details]
Conflicts: [NONE | list with sources]
Proceed? [yes / skip / remove-overlap]
```
Wait for user decision on each item before proceeding.
## Phase 3: Install
### Location Prompt
Always ask before installing:
```
Where should [name] be installed?
1. Global (~/.agentConfig/skills/ or ~/.agentConfig/agents/)
2. Repo-local (.claude/skills./ or .claude/agents./)
```
Use the appropriate subdirectory based on item type (skill vs agent).
### Fetch from GitHub
**Step 1 — Direct download (public repos):**
```bash
# Get tree listing
TREE=$(curl -s "https://api.github.com/repos/{owner}/{repo}/git/trees/{ref}?recursive=1" \
-H "Accept: application/vnd.github.v3+json")
# For skills: download each file in the skill directory
# For agents: download the single .md file
curl -s "https://raw.githubusercontent.com/{owner}/{repo}/{ref}/{filepath}" -o "{local_path}"
```
**Step 2 — Sparse checkout fallback (private repos or download failure):**
```bash
git clone --filter=blob:none --sparse "https://github.com/{owner}/{repo}.git" /tmp/install-tooling-tmpFrom bzellman/earp-kit, MIT.
More from bzellman/earp-kit
- 0.414score
ios-design
bzellman
Design and implement SwiftUI interfaces for <YOUR_APP> iOS app with design system compliance, Apple HIG principles, and visual references.
SkillDFreshness: DClaude Code +9 - 0.411score
ios-simulator-skill
bzellman
21 production-ready scripts for iOS app testing, building, and automation.
SkillDFreshness: DClaude Code +9 - 0.409score
brand-guidelines
bzellman
Applies <YOUR_APP>'s official brand colors and typography to any artifact.
SkillDFreshness: DClaude Code +9 - 0.406score
dead-code-cleanup
bzellman
Safely find and remove dead code from Swift files using an isolated git worktree.
SkillDFreshness: DClaude Code +9 - 0.401score
self-healing-deploy
bzellman
Self-healing merge-to-production pipeline.
SkillDFreshness: DClaude Code +9 - 0.399score
<YOUR_IOS_BUILD>
bzellman
Run <YOUR_APP> iOS builds and simulator tests through FULLSTACK_SCRIPTS/build-ios-formatted.sh and format-build-log.py.
SkillDFreshness: DClaude Code +9 - 0.397score
bug-to-pr
bzellman
End-to-end pipeline from bug report to merged fix PR.
SkillDFreshness: DClaude Code +9 - 0.385score
find-skills
bzellman
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…
SkillDFreshness: DClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Apr 9, 2026 (6 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/bzellman/earp-kit
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from bzellman
- 0.414score
ios-design
bzellman
Design and implement SwiftUI interfaces for <YOUR_APP> iOS app with design system compliance, Apple HIG principles, and visual references.
SkillDFreshness: DClaude Code +9 - 0.411score
ios-simulator-skill
bzellman
21 production-ready scripts for iOS app testing, building, and automation.
SkillDFreshness: DClaude Code +9 - 0.409score
brand-guidelines
bzellman
Applies <YOUR_APP>'s official brand colors and typography to any artifact.
SkillDFreshness: DClaude Code +9 - 0.406score
dead-code-cleanup
bzellman
Safely find and remove dead code from Swift files using an isolated git worktree.
SkillDFreshness: DClaude Code +9 - 0.401score
self-healing-deploy
bzellman
Self-healing merge-to-production pipeline.
SkillDFreshness: DClaude Code +9 - 0.399score
<YOUR_IOS_BUILD>
bzellman
Run <YOUR_APP> iOS builds and simulator tests through FULLSTACK_SCRIPTS/build-ios-formatted.sh and format-build-log.py.
SkillDFreshness: DClaude Code +9
Cite this page
install-tooling skill. Agent Mart. https://agentmart.dev/skills/bzellman-install-tooling (last updated 2026-10-24)