.claude/skills/auto-connect/SKILL.md
Automatically detect installed apps, running services, and logged-in accounts — then connect to them without asking the user. Use when starting a new task to understand what tools are available, when the user mentions an app, or when you need to figure out how to accomplish something without manual user steps.
npx skillsauth add shalevamin/The-_Ultimate_agents auto-connectInstall 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.
Scan the user's environment and automatically connect to all available tools, apps, and services.
# Load environment
ENV_FILE="$(find ~ -name '.env' -path '*/tua-agent/*' 2>/dev/null | head -1)"
[ -f "$ENV_FILE" ] && source "$ENV_FILE" 2>/dev/null || true
echo "=== TUA AGENT — ENVIRONMENT SCAN ==="
echo ""
# ── INSTALLED APPS ──
echo "📱 INSTALLED APPS:"
apps=("Google Chrome" "Safari" "Firefox" "Microsoft Word" "Microsoft Excel" "Microsoft PowerPoint" "Slack" "Visual Studio Code" "Notion" "Figma" "Xcode" "Discord" "Zoom" "Spotify" "1Password" "Obsidian" "Linear" "Things 3" "Bear" "Mailmate" "Airmail" "Spark" "Finder")
for app in "${apps[@]}"; do
if [ -d "/Applications/$app.app" ] || [ -d "$HOME/Applications/$app.app" ]; then
echo " ✅ $app"
fi
done
# ── RUNNING PROCESSES ──
echo ""
echo "🔄 RUNNING:"
pgrep -x "Google Chrome" >/dev/null && echo " ✅ Chrome (active — accounts likely logged in)"
pgrep -x "Safari" >/dev/null && echo " ✅ Safari (active)"
pgrep -x "Slack" >/dev/null && echo " ✅ Slack (active — can send messages)"
pgrep -x "Discord" >/dev/null && echo " ✅ Discord (active)"
pgrep -x "Zoom" >/dev/null && echo " ✅ Zoom (active)"
pgrep -x "Spotify" >/dev/null && echo " ✅ Spotify (active — can control playback)"
pgrep -x "Code" >/dev/null && echo " ✅ VS Code (active)"
# ── CLI TOOLS ──
echo ""
echo "🛠️ CLI TOOLS:"
command -v node &>/dev/null && echo " ✅ Node.js $(node --version)"
command -v python3 &>/dev/null && echo " ✅ Python $(python3 --version 2>&1 | cut -d' ' -f2)"
command -v git &>/dev/null && echo " ✅ Git $(git --version | cut -d' ' -f3)"
command -v docker &>/dev/null && echo " ✅ Docker $(docker --version 2>&1 | cut -d' ' -f3)"
command -v brew &>/dev/null && echo " ✅ Homebrew"
command -v npm &>/dev/null && echo " ✅ npm $(npm --version)"
command -v pip3 &>/dev/null && echo " ✅ pip3"
command -v uv &>/dev/null && echo " ✅ uv (fast Python)"
command -v gh &>/dev/null && echo " ✅ GitHub CLI"
command -v aws &>/dev/null && echo " ✅ AWS CLI"
command -v gcloud &>/dev/null && echo " ✅ Google Cloud CLI"
# ── PYTHON PACKAGES ──
echo ""
echo "🐍 PYTHON PACKAGES:"
python3 -c "import browser_use" 2>/dev/null && echo " ✅ browser-use (AI browser automation)"
python3 -c "import playwright" 2>/dev/null && echo " ✅ playwright (browser control)"
python3 -c "import pyautogui" 2>/dev/null && echo " ✅ pyautogui (GUI automation)"
python3 -c "import anthropic" 2>/dev/null && echo " ✅ anthropic SDK"
python3 -c "import openai" 2>/dev/null && echo " ✅ openai SDK"
python3 -c "import langchain_anthropic" 2>/dev/null && echo " ✅ langchain-anthropic"
python3 -c "import pandas" 2>/dev/null && echo " ✅ pandas (data analysis)"
python3 -c "import PIL" 2>/dev/null && echo " ✅ Pillow (image processing)"
python3 -c "import requests" 2>/dev/null && echo " ✅ requests (HTTP)"
# ── BROWSER TABS ──
echo ""
echo "🌐 CHROME TABS (active sites you're likely logged into):"
osascript 2>/dev/null << 'AS'
tell application "Google Chrome"
if it is running then
set allURLs to {}
repeat with w in windows
repeat with t in tabs of w
set end of allURLs to URL of t
end repeat
end repeat
repeat with u in allURLs
-- Show logged-in platforms
if u contains "google.com" or u contains "docs.google.com" then
log " ✅ Google (Docs/Sheets/Slides available)"
end if
if u contains "notion.so" then
log " ✅ Notion (open)"
end if
if u contains "github.com" then
log " ✅ GitHub (logged in)"
end if
if u contains "linear.app" then
log " ✅ Linear (open)"
end if
if u contains "figma.com" then
log " ✅ Figma (open)"
end if
if u contains "slack.com" then
log " ✅ Slack web (open)"
end if
end repeat
end if
end tell
AS
# ── API KEYS ──
echo ""
echo "🔑 API KEYS:"
[ -n "$ANTHROPIC_API_KEY" ] && echo " ✅ Anthropic API key (configured)" || echo " ❌ ANTHROPIC_API_KEY not set — run setup.sh"
[ -n "$OPENAI_API_KEY" ] && echo " ✅ OpenAI API key"
[ -n "$GOOGLE_API_KEY" ] && echo " ✅ Google API key"
# ── GIT CONFIG ──
echo ""
echo "📋 GIT:"
git config --global user.name 2>/dev/null && echo " ✅ Git user: $(git config --global user.name)"
git config --global user.email 2>/dev/null && echo " ✅ Git email: $(git config --global user.email)"
echo ""
echo "=== SCAN COMPLETE ==="
echo "Tua Agent is now connected to all available tools."
If a tool is needed but not installed:
# Install browser automation
pip3 install --user browser-use playwright langchain-anthropic 2>/dev/null
python3 -m playwright install chromium 2>/dev/null
# Install GUI automation
pip3 install --user pyautogui pillow 2>/dev/null
# Install GitHub CLI
brew install gh 2>/dev/null
# Install Node packages
npm install -g @anthropic-ai/claude-code 2>/dev/null || npx @anthropic-ai/claude-code --help 2>/dev/null
# Create new Word document
osascript << 'AS'
tell application "Microsoft Word"
activate
set newDoc to make new document
set content of text object of newDoc to "DOCUMENT_CONTENT"
save newDoc in (POSIX file "/tmp/tua-document.docx")
end tell
AS
osascript -e 'tell application "Google Chrome" to open location "https://docs.google.com/document/create"'
sleep 3
# Then use keyboard to paste content
osascript << 'AS'
tell application "System Events"
keystroke "DOCUMENT_CONTENT"
end tell
AS
# Send message to a Slack channel
osascript << 'AS'
tell application "Slack"
activate
end tell
AS
# Then use keyboard shortcut to open channel
osascript -e 'tell application "Spotify" to play track "spotify:track:TRACK_ID"'
osascript -e 'tell application "Spotify" to pause'
osascript -e 'tell application "Spotify" to next track'
osascript << 'AS'
tell application "Calendar"
tell calendar "Work"
make new event with properties {summary:"EVENT_TITLE", start date:date "EVENT_DATE", end date:date "EVENT_END"}
end tell
end tell
AS
After scanning, output a capabilities summary:
🤖 TUA AGENT — READY
Available tools:
• Browser: Chrome (logged in as [email protected])
• Documents: Google Docs ✓, Word ✓
• Communication: Slack ✓
• Code: VS Code ✓, Python ✓, Node ✓
• AI: Anthropic ✓ (single API key)
What would you like to build?
development
Use when building cross-platform applications with Flutter 3+ and Dart. Invoke for widget development, Riverpod/Bloc state management, GoRouter navigation, platform-specific implementations, performance optimization.
testing
Use when fine-tuning LLMs, training custom models, or adapting foundation models for specific tasks. Invoke for configuring LoRA/QLoRA adapters, preparing JSONL training datasets, setting hyperparameters for fine-tuning runs, adapter training, transfer learning, finetuning with Hugging Face PEFT, OpenAI fine-tuning, instruction tuning, RLHF, DPO, or quantizing and deploying fine-tuned models. Trigger terms include: LoRA, QLoRA, PEFT, finetuning, fine-tuning, adapter tuning, LLM training, model training, custom model.
tools
Use the Figma MCP server to fetch design context, screenshots, variables, and assets from Figma, and to translate Figma nodes into production code. Trigger when a task involves Figma URLs, node IDs, design-to-code implementation, or Figma MCP setup and troubleshooting.
tools
Translate Figma nodes into production-ready code with 1:1 visual fidelity using the Figma MCP workflow (design context, screenshots, assets, and project-convention translation). Trigger when the user provides Figma URLs or node IDs, or asks to implement designs or components that must match Figma specs. Requires a working Figma MCP server connection.