The Agent Skills Directory

[COMMAND_EXECUTION]: Includes setup scripts (scripts/setup-codex-hook.sh, scripts/setup-gemini-hook.sh) that modify user-level configuration files such as ~/.codex/config.toml and ~/.gemini/settings.json to install agent hooks and instructions.
[COMMAND_EXECUTION]: Creates and makes executable a bash hook script (~/.gemini/hooks/ralph-ooo-check.sh) to enable automated loop continuation in the Gemini CLI environment.
[COMMAND_EXECUTION]: The scripts/ooo-state.sh script manages a local JSON state file in .omc/state/ralph-ooo-state.json to persist session progress and iteration counts.
[EXTERNAL_DOWNLOADS]: Documentation references the author's and the project's GitHub repositories (akillness/oh-my-skills, Q00/ouroboros) for installation and updates via npx or native plugins.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted user input through ooo interview and ooo ralph (Ingestion points) without defining boundary markers or sanitization logic (Sanitization: absent, Boundary markers: absent). The agent environment provides capabilities including Bash, Write, and Agent tools (Capability inventory).

ralph