The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from Figma layer names and text nodes which could contain malicious instructions.
Ingestion points: Node data is retrieved via scan_nodes_by_types, get_node, and get_selection tools as described in SKILL.md.
Boundary markers: The instructions require the agent to present a preview table to the user for confirmation before applying any changes.
Capability inventory: The skill possesses the rename_node capability to modify data in the Figma environment.
Sanitization: There is no mention of sanitizing or escaping the content of layer names or text nodes before they are processed by the agent.
[COMMAND_EXECUTION]: The skill utilizes the figma-mcp-rust toolset to perform operations on the Figma document, including reading node properties and executing rename commands.

bulk-rename