skills/greeting/SKILL.md
Greet all installed tools and generate a health report
npx skillsauth add michelabboud/claude-code-helper greetingInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Survey all installed tools — MCP servers, agents, and skills — by sending hello to each one. Generate a formatted health report showing what's installed and responding.
/greeting → Quick hello to all MCP servers + list agents & skills + summary report
/greeting ID → Verbose hello (full profile) from all servers + complete report
When invoked, follow the steps below based on the argument:
Step 1 — Greet MCP Servers
Call hello {} on each of the following MCP servers by using their hello tool. If a tool is unavailable or throws an error, mark it ❌ Offline.
| Server | Tool to call |
|--------|-------------|
| code-review-mcp | mcp__code-review__hello with {} |
| testing-mcp | mcp__testing__hello with {} |
| design-system-mcp | mcp__design-system__hello with {} |
| api-specialist-mcp | mcp__api-specialist__hello with {} |
| uiux-review-mcp | mcp__uiux-review__hello with {} |
| database-operations | mcp__database-operations__hello with {} |
| dependency-management | mcp__dependency-management__hello with {} |
| cicd-pipeline | mcp__cicd-pipeline__hello with {} |
| n8n-automation | mcp__n8n-automation__hello with {} |
| rag-mcp | mcp__rag__hello with {} |
| project-oversight-mcp | mcp__claude_ai_Vercel__search_vercel_documentation is NOT this server — skip. Use the hello tool if available. |
Show each server's greeting response, prefixed with ✅ (responded) or ❌ (failed).
Step 2 — List Installed Agents
Use the Glob tool to find all agent files in ~/.claude/agents/:
~/.claude/agents/**/*.md and ~/.claude/agents/**/*.jsonFor each file found:
.md) or top-level JSON fields (.json) to extract: name, description, color, and for JSON: visual.emojicolor to emoji: green=🟢, blue=🔵, purple=🟣, red=🔴, orange=🟠, yellow=🟡, cyan=🩵, default=⚪visual.emoji if present instead of the color square[emoji] **name** — descriptionGroup output as:
## Installed Agents (X found)
🔵 **code-reviewer** — Reviews code for bugs, security issues, and best practices
🟣 **architect** — System design and architecture planning
...
Step 3 — List Installed Skills
Use the Glob tool to find all skills in ~/.claude/skills/:
~/.claude/skills/**/SKILL.md and ~/.claude/skills/**/*.md (excluding SKILL.md if already found)For each skill found:
name, version, description📘 **name** v[version] — descriptionGroup output as:
## Installed Skills (X found)
📘 **greeting** v1.0.0 — Greet all installed tools and generate a health report
📘 **pm-dashboard** v1.0.0 — Manage the Project Manager dashboard
...
Step 4 — Generate Report
Output a formatted summary block:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📊 Greeting Report — [current date/time]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
MCP Servers X / 10 online
Agents X installed
Skills X installed
[If any servers were offline:]
⚠️ Offline servers: server-name, server-name
→ Run: cd mcp-servers/[name] && npm run build
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
IDSame as above, but:
hello {"verbose": true} on each MCP server instead of {}The report format is the same but titled Full Greeting Report.
helloRespond with:
👋 Hello! I'm Greeting v1.0.0. I survey all installed MCP servers, agents, and skills — and generate a health report. Use
/greeting hello IDfor the full guide.
hello IDRespond with complete skill information:
/greeting [ID]ID — Verbose hello (full profile) from all servers + complete reporthello — Quick greeting + availability checkhello ID — This full profiledevelopment
Score a coding task by complexity (1-10) and recommend the right model (haiku/sonnet/opus) before invoking a language agent. Holds the per-language rubric for all language/framework experts.
tools
When the user asks about [TRIGGER WORDS], wants to [ACTION], or needs help with [TOPIC], use this skill to provide [CAPABILITY]
tools
Check if your claude-code-helper installation is up to date and apply updates. Reads the local manifest and compares against the latest component-versions index on GitHub. Supports checking all components, a single component by name, and applying updates with automatic backup.
testing
Comprehensive testing skill covering TDD, E2E, BDD, contract testing, mutation testing, and visual regression. Use when writing tests, designing test strategy, adding test coverage, fixing flaky tests, mocking services, setting up testing frameworks, or any testing task. Triggers on 'write tests', 'add test coverage', 'test strategy', 'fix flaky test', 'mock', 'E2E test', 'unit test', 'integration test'.