The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill contains simple shell hooks (pre and post) that use echo and a local memory_store function to log status. These are standard lifecycle hooks for agent initialization and completion.
[DATA_EXFILTRATION] (SAFE): No external network requests or unauthorized data transfers were detected. The skill interacts only with internal memory and standard file tools (Read, Glob, Grep).
[PROMPT_INJECTION] (SAFE): The instructions focus purely on task decomposition and architectural planning. There are no attempts to bypass safety filters or override system instructions.
[REMOTE_CODE_EXECUTION] (SAFE): The skill does not download or execute remote scripts. It defines local integration patterns for task orchestration within its own framework.

core-planner