The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use shell tools such as find, ls, grep, and sort to locate briefing files within the docs/rick directory. It incorporates a safeguard by recommending a regex validation pattern (^[a-z0-9][a-z0-9-]*$) to sanitize folder name arguments before they are used in shell commands.
[PROMPT_INJECTION]: The skill identifies an indirect prompt injection surface as it processes external data from briefing files to determine the session state. Specifically, it scans for a trigger line (You are Rick. You know what to do.) that signifies a prior persona mode.
Ingestion points: Briefing files located at docs/rick/<folder>/saves/<timestamp>.md.
Boundary markers: No specific delimiters are used to wrap the ingested content.
Capability inventory: The agent performs file reads and shell-based file discovery.
Sanitization: The skill mitigates the risk of unauthorized persona activation by explicitly instructing the agent that the detected signal must not trigger an automatic behavioral change, requiring manual user intervention instead.

rick-respawn