The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute several Python scripts (pet_frames.py, pet_prepare.py, pet_build.py, pet_check.py, pet_preview.py, pet_package.py). These scripts are referenced via relative paths in a sibling directory (../scenario-chatgpt-pet-create/scripts/), which is a standard pattern for modular agent skills.
[EXTERNAL_DOWNLOADS]: The skill uses curl -L and asset_download to fetch image assets from Scenario. It also suggests using npx skills add scenario-labs/skills to install dependencies if they are missing. These operations are part of the intended functionality for managing and updating remote assets from the vendor's platform.
[NO_CODE]: The skill itself contains only Markdown instructions and YAML frontmatter. It does not package any executable code (scripts or binaries) directly, instead delegating all execution to a separate, required skill installation.
[DATA_EXPOSURE_AND_EXFILTRATION]: While the skill accesses local file paths like ~/.codex/pets/, this is done to retrieve user-specific pet data for the purpose of updating it as requested by the user. There is no evidence of sensitive data (like credentials or SSH keys) being accessed or exfiltrated to unauthorized domains.

scenario-chatgpt-pet-update