The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts (knowledge-binding.py and health.py) using the Bash tool to audit the structure of knowledge bases. These scripts are located within the plugin's own environment and the execution is restricted to the specific diagnostic logic described.
[DATA_EXFILTRATION]: No network activity or external data transmission was identified. The skill is designed to operate as a read-only structural validator that outputs results directly to the agent's stdout for user consumption.
[PROMPT_INJECTION]: The instructions provide operational guidance for running health checks and do not contain patterns for bypassing safety filters, system prompt extraction, or role-play based jailbreaks.
[DYNAMIC_EXECUTION]: The skill dynamically resolves the path for the health.py engine using a local helper script (resolve-wiki-scripts.sh). This is used to locate internal vendored dependencies within the plugin's filesystem and does not involve executing code from untrusted external sources.

knowledge-health