The Agent Skills Directory

[SAFE]: The skill is designed for information extraction and local storage. It does not utilize network operations or execute external code.
[DATA_EXPOSURE]: Access is restricted to a specific local directory (./writing-workspace/materials/) for managing research materials. No access to sensitive system paths or credentials was detected.
[COMMAND_EXECUTION]: The skill does not invoke shell commands or system-level processes.
[INDIRECT_PROMPT_INJECTION]: The skill processes external text provided by users. While this presents an ingestion surface for untrusted data, the skill's capabilities are limited to structured data extraction and file writing within a restricted workspace. The instructions specifically require proper JSON escaping to prevent data corruption or injection during the writing process.
Ingestion points: User-provided articles or text inputs.
Boundary markers: Not explicitly defined in the prompt template, but processing is scoped to extraction.
Capability inventory: File read/write and directory creation within ./writing-workspace/materials/.
Sanitization: Explicit instructions are included to escape special characters like double quotes and backslashes during JSON serialization.

material-ingest