The Agent Skills Directory

[DATA_EXPOSURE]: The skill accesses codebase information using a specialized explore agent and manages state/plans within the .omc/ directory. This is consistent with its stated purpose as a planning tool and does not involve unauthorized data access.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the codebase and existing plan files. While this presents a surface for indirect prompt injection, the skill includes human-in-the-loop checkpoints (in --interactive mode) and outputs structured markdown plans rather than executing code directly.
[COMMAND_EXECUTION]: The skill invokes other internal tools and agents (e.g., ralph, team, architect, critic) within the oh-my-claudecode namespace for specialized tasks. These are restricted to the agent's internal ecosystem and are managed via a state lifecycle to prevent unexpected behavior.

omc-plan