The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/http_benchmark.py uses the subprocess module to run external performance testing binaries (wrk or ab).
Evidence: The functions run_wrk and run_ab construct command lists which are passed to subprocess.run to execute the load tests.
Risk Assessment: The script uses the list-based invocation method for subprocess.run rather than shell=True, which effectively prevents shell injection attacks. Furthermore, the script includes a validate_url function that ensures target URLs use only the http or https protocols and contain valid hostnames.
[SAFE]: The skill performs its stated purpose of load testing and performance analysis. It does not contain any obfuscated code, unauthorized data exfiltration patterns, or attempts to gain persistence. The use of external tools is documented in the SKILL.md file along with clear instructions for the user.

http-load-tester