The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is designed to operate through a command-line utility called opencli. It executes shell commands to list available search sources (opencli list), inspect command help (opencli <site> -h), and perform the actual searches. This is the intended primary function of the skill.
[SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were detected. The skill demonstrates defensive design by requiring the agent to perform a 'mandatory pre-check' of tool commands before execution to prevent errors from outdated documentation (document drift). Additionally, it enforces a budget/limit on the number of calls per site to prevent runaway processes or platform abuse.

smart-search