skills/jeremylongshore/claude-code-plugins-plus-skills/analyzing-system-throughput/Gen Agent Trust Hub
analyzing-system-throughput
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is focused on performance monitoring and optimization. Detailed inspection of the bundled scripts confirms they are intended for local data analysis. The file
identify_bottlenecks.shcontains Python code despite its shell extension, which appears to be a naming convention for this toolset. - [COMMAND_EXECUTION]: The skill executes local Python scripts to automate metric collection. These scripts use standard libraries and avoid dangerous practices such as executing arbitrary strings or downloading external payloads.
- [DATA_EXPOSURE]: The scripts perform recursive directory traversal to gather file metadata (size and counts). This metadata is used locally for reporting and is not transmitted to external servers. No access to sensitive system files was observed.
- [PROMPT_INJECTION]: The skill ingests external metric data for analysis. While this is a surface for indirect prompt injection, the risk is negligible as the scripts do not interpret the ingested data as executable instructions.
Audit Metadata