The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands including md5 and rg (ripgrep) using inputs derived from user arguments. The instruction md5 -q <input_file> directly interpolates a user-supplied file path or argument into a shell command, which allows for command injection if the input contains shell metacharacters.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from various sources and possesses significant capabilities, meeting the criteria for a vulnerable injection surface.
Ingestion points: Content from external URLs (YouTube, Instagram, recipe sites) and local archive metadata (.yaml files) stored in the /Users/mtm/pdev/taylormonacelli/grouchygiraffe/data directory.
Boundary markers: None; there are no instructions to use delimiters or to disregard instructions embedded in the external or local data sources.
Capability inventory: Execution of shell commands, invocation of auxiliary skills (recipe-cleanup, social-to-imgur, grouchygiraffe-recipe-lookup), and writing processed recipes to a local vault.
Sanitization: The skill does not perform any sanitization or validation of the data retrieved from external sources before using it to influence downstream tool calls or writing it to the filesystem.

grouchygiraffe-ingest