The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified in the analyzed files.
[DATA_EXFILTRATION]: No network operations or attempts to access sensitive system files (such as credentials or SSH keys) were detected. All scripts process data locally.
[REMOTE_CODE_EXECUTION]: The skill does not use dangerous functions like eval() or exec(), nor does it attempt to download and execute scripts from the internet.
[COMMAND_EXECUTION]: While the documentation provides examples of CLI usage, the scripts themselves do not programmatically execute shell commands or use the subprocess module.
[PROMPT_INJECTION]: The instructions and documentation do not contain patterns intended to bypass AI safety guidelines or override system behavior.
[INDIRECT_PROMPT_INJECTION]: The skill includes an input format detector that parses user-provided text, YAML, and JSON. While this is an ingestion point for untrusted data, the parsed output is used strictly for mathematical scoring and markdown report generation, which does not pose a high security risk.

tech-stack-evaluator