The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it processes untrusted external data such as meeting transcripts, RFPs, and research papers. \n
Ingestion points: Processes diverse inputs including idea descriptions, RFPs, internal briefs, research dossiers, and meeting transcripts as defined in the Intake phase (SKILL.md). \n
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' directives for instructions embedded within the processed data. \n
Capability inventory: The skill has access to powerful tools including web_search, tool_search (accessing Drive and Gmail), and local file system write access for generating .md, .docx, and .pptx files (SKILL.md). \n
Sanitization: No specific input sanitization or validation logic is defined in the workflow instructions.

proposal-generator