The Agent Skills Directory

[SAFE]: No malicious code, obfuscation, or unauthorized network operations were detected within the skill files.
[COMMAND_EXECUTION]: The skill includes a bash wrapper ('run.sh') used to configure the local environment and execute the entity extraction script. This activity is restricted to the local execution of the tool's core logic.
[PROMPT_INJECTION]: While the skill accepts user-provided text as input, its primary function is to resolve that text against grounded data, which acts as a security control to detect and mitigate potential fabrications or adversarial inputs.

extract-entities