The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted input from user-provided business ideas, inspiration sources (such as tweets or external chats), and existing project memory files (project_*.md). This data is used to score ideas and is saved into permanent archives, which could be exploited if the input contains malicious instructions designed to influence the agent's logic or the content of the generated reports.
Ingestion points: User-provided business ideas, reasoning context, external project memory files, and local portfolio configuration documents.
Boundary markers: The instructions do not define specific delimiters or guidelines to distinguish between system instructions and interpolated untrusted data during processing.
Capability inventory: The skill has the ability to write to the local file system (archiving to ~/.config/makerskills/) and invoke other functional skills such as deep-research and /domain.
Sanitization: There is no evidence of explicit input validation or sanitization before the data is processed or stored in the archive.
[COMMAND_EXECUTION]: The skill directs the agent to perform file system operations, specifically moving files from an old directory structure to a new archive location during its migration step. This utilizes the agent's file management capabilities to organize and maintain user data.

business-brainstorm