analyze
Run a comprehensive multi-agent crypto analysis with phased execution. Usage: /analyze BTC or /analyze ETH SOL
First-time setup for Crypto Trading Desk. Detects your environment, installs dependencies, and verifies everything works. Run this once after installing the plugin.
$ npx -y skills add hugoguerrap/crypto-claude-desk --skill setup --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/setupContext preview
The summary Claude sees to decide when to auto-load this skill.
First-time setup for Crypto Trading Desk. Detects your environment, installs dependencies, and verifies everything works. Run this once after installing the plugin.
name: setup description: First-time setup for Crypto Trading Desk. Detects your environment, installs dependencies, and verifies everything works. Run this once after installing the plugin. user-invocable: true
Run environment detection and dependency setup. This skill makes the plugin work on any platform.
Run these checks in parallel using Bash:
Based on detected OS:
**macOS / Linux:**
curl -LsSf https://astral.sh/uv/install.sh | sh
**Windows (PowerShell):**
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
After installing, verify `uv` is accessible. If it's installed to `~/.local/bin/` but not in PATH (common on macOS GUI apps), offer to create a symlink:
sudo ln -sf ~/.local/bin/uv /usr/local/bin/uv
If the user declines sudo, tell them to restart their terminal and try again.
uv sync --project <plugin_root>
This creates `.venv/` and installs all packages from `uv.lock` (fastmcp, ccxt, numpy, pandas, requests).
If `uv sync` fails, try without `--frozen`:
uv sync --project <plugin_root>
mkdir -p data/trades data/reports data/logs data/db .claude/agent-memory/portfolio-manager
The SQLite database (`data/db/learning.db`) is created automatically by the crypto-learning-db MCP server on first use. No manual initialization needed.
Test that one MCP server can start by running:
uv run --frozen --project <plugin_root> python -c "import fastmcp; import ccxt; import numpy; import pandas; print('OK')"If this fails, diagnose and fix. Common issues:
Read `~/.claude/settings.json` and check if MCP tool permissions are allowed.
If not, show the user what to add and offer to configure it.
Present a clear status report:
Crypto Trading Desk - Setup Complete
OS: macOS 15.3 (arm64)
Python: 3.12.8
uv: 0.6.13
Dependencies: 5/5 installed
MCP Servers: 7/7 ready
Data dirs: Created
Database: SQLite ready ($20,000 paper trading)
Ready to use:
/crypto-trading-desk:quick BTC (fast market check)
/crypto-trading-desk:analyze ETH (full 5-agent analysis)If anything failed, show exactly what went wrong and how to fix it.
I used to spend weeks building multi-agent systems with LangGraph, CrewAI, and AutoGen. Hundreds of lines of Python orchestration code, custom state machines, fragile message passing between agents.
Repo: hugoguerrap/crypto-claude-desk
Run a comprehensive multi-agent crypto analysis with phased execution. Usage: /analyze BTC or /analyze ETH SOL
Close an open trade and run post-mortem analysis. Usage: /close-trade trade_001 or /close-trade trade_001 at 98500
Extend the system by creating new MCP servers, agents, or skills. Usage: /create a DeFi protocol tracker or /create an agent for macro analysis
Autonomous monitoring loop. Checks open trades against SL/TP levels, closes trades that hit targets, evaluates expired predictions, and generates periodic…
View current portfolio status, open trades, and performance stats. Usage: /portfolio