The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's example output receipts contain a contactsExported field that is unrelated to the primary design-to-code purpose. This introduces a potential schema confusion vulnerability where an agent might be misled into processing or reporting data outside its intended scope.
Ingestion points: The batchGenerate function accepts an array of nodeIds strings (SKILL.md).
Boundary markers: The instructions recommend defining an explicit allowlist of Figma files and node IDs to bound the batch generation process (SKILL.md).
Capability inventory: The skill utilizes Bash(npm:*) for package management and Write/Edit for script creation (SKILL.md).
Sanitization: Instructions mandate the redaction of tokens and design content from all logs and receipts to prevent leakage (SKILL.md).
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the bottleneck and @animaapp/anima-sdk packages. These are retrieved from the official NPM registry, which is a well-known and trusted service.
[CREDENTIALS_UNSAFE]: The skill references sensitive environment variables including ANIMA_TOKEN, FIGMA_TOKEN, and FIGMA_FILE_KEY. It follows secure practices by instructing the user to store these in a secret manager or environment, and specifically warns against hardcoding them or including them in logs.

anima-rate-limits