Skip to content

SKILL.md

Latest commit

 

History

History
169 lines (128 loc) · 4.44 KB

SKILL.md

File metadata and controls

169 lines (128 loc) · 4.44 KB
name local-code-interpreter
description Execute Python code and shell commands locally on macOS/Linux. No API key, no cloud service — runs directly on your machine.
metadata
openclaw
os emoji user-invocable homepage
darwin
linux
🐍
true

Local Code Interpreter

Run Python code and shell commands directly on your local machine. No cloud, no API keys, no cold starts — uses the Python already installed on your system.

When to Use

Use this skill whenever you need to actually execute code to get a result:

Task Use This?
Data analysis with pandas/numpy ✅ Yes
Generate charts and visualizations ✅ Yes
Process files (CSV, Excel, JSON, PDF) ✅ Yes
Machine learning with scikit-learn ✅ Yes
Math / statistics calculations ✅ Yes
Explaining code or algorithms ❌ No — respond with text
Formatting code examples ❌ No — use markdown code blocks
Simple mental math ❌ No — just answer directly

How to Execute Code

Single-line code

python3 {baseDir}/scripts/run.py --code "print('hello')"

Multi-line code (recommended)

Write code to a temp file, then execute:

cat > /tmp/ci_script.py << 'PYEOF'
import pandas as pd
import numpy as np

data = {'month': ['Jan','Feb','Mar'], 'revenue': [12, 19, 15]}
df = pd.DataFrame(data)
print(df.describe())
PYEOF
python3 /tmp/ci_script.py

Execute a file directly

python3 {baseDir}/scripts/run.py --file /path/to/script.py

Shell commands

Use the built-in execute_command tool directly — no wrapper needed:

pip3 install pandas matplotlib
ls -la

Rules

  1. Non-interactive matplotlib backend — always add before pyplot import:

    import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

  2. Chinese text in charts — set font to avoid missing glyphs:

    plt.rcParams['font.family'] = 'PingFang HK'   # macOS
# plt.rcParams['font.family'] = 'WenQuanYi Zen Hei'  # Linux

  3. Use print() for output — stdout is the only output channel.

  4. Save charts with savefig(), never plt.show():

    plt.savefig('/tmp/output.png', dpi=150, bbox_inches='tight')

  5. Default output directory is /tmp/ — specify a different path explicitly if needed.

  6. Install missing packages first:

    pip3 install <package>

Pre-installed Libraries (common)

Install any missing ones with pip3 install <name>.

Category Libraries
Data Analysis pandas, numpy
Visualization matplotlib, plotly, seaborn
Machine Learning scikit-learn, xgboost
Excel / Office openpyxl, python-docx, python-pptx
PDF PyPDF2, reportlab, pdfplumber
Image Pillow, opencv-python
HTTP / Scraping requests, httpx, beautifulsoup4
Math scipy, sympy
Utilities rich, pydantic, tqdm

Example Workflows

Data Analysis

import pandas as pd
import numpy as np

df = pd.read_csv('/path/to/data.csv')
print("Shape:", df.shape)
print("\nMissing values:")
print(df.isnull().sum())
print("\nStatistics:")
print(df.describe())

Visualization (Chinese labels)

import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import numpy as np

plt.rcParams['font.family'] = 'PingFang HK'

months = ['1月', '2月', '3月', '4月', '5月', '6月']
values = [12, 19, 15, 25, 22, 30]

plt.figure(figsize=(8, 4))
plt.plot(months, values, 'o-', color='#2196F3', linewidth=2)
plt.title('上半年营收趋势（万元）')
plt.xlabel('月份')
plt.ylabel('营收（万元）')
plt.grid(True, alpha=0.3)
plt.tight_layout()
plt.savefig('/tmp/chart.png', dpi=150, bbox_inches='tight')
print('Chart saved to /tmp/chart.png')

Install and Use New Package

pip3 install yfinance
import yfinance as yf
ticker = yf.Ticker("AAPL")
info = ticker.info
print(f"Apple stock price: ${info.get('currentPrice', 'N/A')}")

Environment

  • Python: 3.13 (macOS) or system Python 3.x (Linux)
  • Shell: zsh (macOS) / bash (Linux)
  • Execution: local process, full filesystem access
  • Networking: full internet access via requests, curl, etc.
  • Session state: variables do NOT persist between separate calls — use files to pass data across steps