The Agent Skills Directory

[SAFE]: The skill serves as a high-quality technical manual for the EffectCraft application. It provides verified functionality descriptions and implementation evidence sourced directly from the application's repository.- [SAFE]: Security limitations and risks inherent to the tool (such as an unauthenticated localhost control port) are transparently documented with accompanying mitigation strategies, such as using root-restricted automation paths.- [INDIRECT_PROMPT_INJECTION]: The skill documents ingestion points for external data including project files (JSON/XML), data footage (CSV/TSV), and Lottie animations. To mitigate potential injection risks from untrusted data, the skill instructs the agent to use built-in validation commands and environment-level security flags like --automation-read-root to constrain the application's reach.- [DYNAMIC_EXECUTION]: The application's support for WebAssembly plugins and JavaScript expressions is documented. The skill outlines the security model for these features, noting that scripts are confined to the project folder and cannot access the network unless explicitly permitted by user settings.

effectcraft