CtrlK
BlogDocsLog inGet started
Tessl Logo

reliable-script-execution

Execute Python scripts reliably using file-first approach instead of heredoc

60

Quality

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./benchmarks/gdpval/skills/reliable-script-execution/SKILL.md
SKILL.md
Quality
Evals
Security

Reliable Script Execution

When executing Python code via shell commands, avoid inline heredoc execution which can fail unpredictably with 'unknown error'. Use this two-step file-first approach for more reliable script execution.

Problem

Direct heredoc Python execution like:

python3 << 'EOF'
# complex code here
EOF

Can fail with 'unknown error', especially when:

  • Script contains multiple lines or complex logic
  • Special characters or quotes are present
  • Working directory context matters

Solution: File-First Approach

Step 1: Write Python Script to File

Use write_file to save your Python code to a .py file:

write_file(path="./temp_script.py", content="""
import json

data = {"key": "value"}
print(json.dumps(data))
""")

Step 2: Execute via Shell

Use run_shell with explicit working directory:

run_shell(command="python3 ./temp_script.py", timeout=60)

Step 3: Clean Up (Optional)

Remove temporary files after execution:

run_shell(command="rm ./temp_script.py")

Complete Example

Task: Generate a JSON report with calculations

# Step 1: Write the script
write_file(
    path="./generate_report.py",
    content="""
import json
from datetime import datetime

revenue = 500000.00
expenses = 379577.06
net_income = revenue - expenses

report = {
    "generated": datetime.now().isoformat(),
    "revenue": revenue,
    "expenses": expenses,
    "net_income": net_income
}

print(json.dumps(report, indent=2))
"""
)

# Step 2: Execute
run_shell(command="python3 ./generate_report.py", timeout=60)

# Step 3: Clean up
run_shell(command="rm ./generate_report.py")

Best Practices

  1. Use descriptive filenames: Name scripts according to their purpose (e.g., calculate_pnl.py, transform_data.py)

  2. Set appropriate timeouts: For data processing scripts, use longer timeouts (60-300 seconds)

  3. Specify working directory: If the script depends on relative paths, include cd /path && python3 script.py

  4. Handle errors gracefully: Check shell output for errors and retry if needed

  5. Clean up temporary files: Don't leave .py files cluttering the workspace unless they need to persist

When to Use This Pattern

  • Executing multi-line Python code via shell
  • Scripts with complex string handling or special characters
  • When heredoc execution has failed previously
  • Any scenario where reliability matters more than brevity

Anti-Pattern to Avoid

Do NOT rely on heredoc for production-critical scripts:

# Unreliable - may fail with 'unknown error'
python3 << 'EOF'
# Your code here
EOF

Use file-first approach instead for consistent, debuggable execution.

Repository
HKUDS/OpenSpace
Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.