The Agent Skills Directory

[SAFE]: The skill's instructions are focused entirely on academic writing, research methods, and manuscript synthesis. There are no patterns of prompt injection, obfuscation, or persistence mechanisms detected.- [DATA_EXFILTRATION]: The skill requests access to read project files and perform web searches for journal guidelines and writing standards. This is consistent with its role as a scholarly writing coach and does not involve unauthorized access to sensitive system paths or credentials.- [PROMPT_INJECTION]: While the skill ingests external data from web searches and user-provided manuscripts (potential indirect prompt injection surface), its instructions strictly enforce academic structure and logical reasoning, which act as constraints. The ingestion of this data is necessary for its primary purpose of synthesising literature and providing writing feedback.- [COMMAND_EXECUTION]: No suspicious command execution or shell injection patterns were identified. The agent's workflow involves reading and writing markdown files, which are standard operations for generating deliverables.

alterlab-rma-academic-writer