The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill is designed to ingest and process untrusted data from multiple external sources, including customer support tickets (references/customer-support.md), user research interviews, and hiring transcripts (references/hr.md).
Boundary markers: The templates for processing this data do not consistently implement boundary markers or explicit instructions for the model to disregard embedded commands in the external text.
Capability inventory: The skill has access to powerful tools including Bash, Write, and Edit (as defined in SKILL.md), which are used to execute data collection scripts and update documentation.
Sanitization: The instructions for research and data analysis do not specify sanitization or validation routines for external data before it is interpolated into shell commands or file operations in references/competitive-mapping.md and references/trend-analysis.md.

business-ops