The Agent Skills Directory

[SAFE]: The skill performs legitimate text processing tasks on local research documents. Its primary functions include thematic clustering, quote extraction, and report generation within the .chalk/ directory.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it reads and processes external data (interview notes, transcripts). This is evaluated as safe in this context as the skill lacks network access to exfiltrate data and its filesystem operations are confined to research documentation. Ingestion points: Qualitative research files provided via $ARGUMENTS. Boundary markers: The instructions do not specify delimiters to separate untrusted data from the system prompt. Capability inventory: Read, Glob, Grep, and Write tools used to manage research documentation. Sanitization: No specific sanitization or filtering of research content is performed.

synthesize-research