The Agent Skills Directory

[SAFE]: The skill implements lookup functionality by executing cat and jq on its internal index files (pipeline-phases-index.json and tools-index.json). This is a legitimate and restricted use of shell commands to provide data to the agent.
[EXTERNAL_DOWNLOADS]: The skill includes documentation for installing various industry-standard security tools. For instance, it references the official installation script for Anchore's Grype from their GitHub repository. These references are provided as informational guidance for the user and are not executed by the skill itself.
[COMMAND_EXECUTION]: Command execution is limited to parsing local JSON data. The skill instructions guide the agent to perform specific, safe queries against its own data files, with no evidence of unsafe user input interpolation or arbitrary command execution.
[DATA_EXFILTRATION]: No network operations or sensitive file access patterns were detected. The skill only interacts with its own bundled resource files.

devsecops-lookup