The Agent Skills Directory

[SAFE]: The skill follows secure design principles for AI agents. It uses a state-managed 'ledger' system for session persistence within the Penpot file metadata, which is a legitimate and safe practice for maintaining state across tool calls.
[SAFE]: The skill manages remote resources safely. The image upload functionality in scripts/applyDeckEffects.js is restricted to user-supplied HTTPS URLs and is handled by the platform's native uploadMediaUrl method. It does not attempt to execute remote code or scripts.
[SAFE]: No obfuscation, prompt injection, or credential harvesting patterns were found. The skill uses static, well-documented scripts for all operations and includes extensive anti-rationalization rules to prevent the agent from deviating from its intended design purpose.

penpot-build-deck