The Agent Skills Directory

[SAFE]: No security issues were detected. The skill contains instructional content and code examples for interacting with the Figma Slides API via the use_figma tool.
[PROMPT_INJECTION]: Instructions provided are functional and aimed at ensuring correct tool usage and adherence to design principles. No attempts to bypass safety filters or override system instructions were found.
[DATA_EXFILTRATION]: The skill interacts with Figma file content as part of its primary function. There are no patterns suggesting data is sent to unauthorized external domains or that sensitive local credentials (e.g., .ssh, .aws) are accessed.
[REMOTE_CODE_EXECUTION]: The code snippets are intended for execution within the Figma Plugin API sandbox via the provided MCP tool. No instances of downloading and executing untrusted remote scripts (e.g., curl | bash) were identified.
[INDIRECT_PROMPT_INJECTION]: The skill includes instructions for reading data from Figma files (e.g., slide text, speaker notes). This represents an indirect injection surface if the processed file content contains malicious instructions.
Ingestion points: Reading speakerNotes in references/slide-properties.md and extracting text content in SKILL.md via findAllWithCriteria.
Boundary markers: None provided in the reference code for handling returned data.
Capability inventory: Uses use_figma for read/write and upload_assets for file uploads.
Sanitization: None specified for external content processing.

figma-use-slides