<< All versions
Skill v1.0.2
currentAutomated scan100/100hkuds/openspace/fallback-python-execution
1 files
──Details
PublishedApril 26, 2026 at 05:08 PM
Content Hashsha256:20bf505b1fd30e49...
Git SHAd1e367d0ed47
Bump Typepatch
──Files
Files (1 file, 2.7 KB)
SKILL.md2.7 KBactive
SKILL.md · 101 lines · 2.7 KB
version: "1.0.2" name: fallback-python-execution description: Reliable Python execution workflow when execute_code_sandbox or shell_agent fail
Fallback Python Execution Pattern
When to Use
Use this pattern when:
execute_code_sandboxreturns unknown errors or fails repeatedlyshell_agentcannot successfully execute Python code- You need to create files (spreadsheets, documents, data files) via Python
- Direct delegated approaches prove unreliable in the current environment
Core Technique
Instead of delegating Python execution to agents, use this two-step inline approach:
- Write Python code to a
.pyfile usingwrite_file - Execute the file using
run_shellwithpython <script.py>
Step-by-Step Instructions
Step 1: Write Python Code to File
Use write_file to create a Python script with all necessary code inline:
write_filepath: /path/to/script.pycontent: |import pandas as pd# Your complete Python code heredf = pd.DataFrame({...})df.to_excel('output.xlsx', index=False)
Step 2: Execute via run_shell
Run the script directly:
run_shellcommand: python /path/to/script.py
Step 3: Verify and Clean Up
- Check the output for success/errors
- Verify the expected files were created
- Optionally remove the temporary script if no longer needed
Why This Works
This approach is more reliable because:
- Avoids agent interpretation layers that can introduce errors
- Provides direct control over execution environment
- Gives clear error output for debugging
- Bypasses sandbox delegation issues
Example: Excel File Creation
yaml
# Step 1: Write the scriptwrite_file:path: create_report.pycontent: |import pandas as pdfrom openpyxl import Workbook# Create datadata = {'Column1': [1, 2, 3], 'Column2': ['A', 'B', 'C']}df = pd.DataFrame(data)# Save to Exceldf.to_excel('report.xlsx', index=False)print('Excel file created successfully')# Step 2: Executerun_shell:command: python create_report.py
Tips
- Include error handling in your Python code for better debugging
- Use absolute paths when possible to avoid working directory issues
- Add print statements to track execution progress
- Keep scripts self-contained with all imports at the top
- For complex tasks, break into multiple scripts if needed
Troubleshooting
| Issue | Solution | |
|---|---|---|
| Module not found | Add pip install commands before python command | |
| Permission errors | Check file paths are writable | |
| Script not found | Use absolute path or cd to directory first | |
| Output not created | Check for Python errors in run_shell output |