The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a bash script (scripts/idea-refine.sh) to initialize a directory for storing ideas. The script uses the mkdir -p command to create a docs/ideas directory if it does not already exist.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input from $ARGUMENTS and local project files using tools like Glob, Grep, and Read.
Ingestion points: Data enters the agent context via user-supplied arguments and automated scanning of the local codebase as described in SKILL.md.
Boundary markers: There are no explicit instructions or delimiters defined in SKILL.md to separate external data from system instructions.
Capability inventory: The agent can execute a local bash script (scripts/idea-refine.sh) which uses mkdir, read files, search project contents, and write refined concepts to markdown files in the docs/ideas/ directory (SKILL.md).
Sanitization: The skill instructions do not specify sanitization, validation, or filtering logic for the content it ingests from external project files or user input.

idea-refine