The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for the agent to construct and execute shell commands using the islandiguana and ripgrep binaries. There is a risk of command injection if the agent interpolates unsanitized user-provided search terms or values into the command arguments, potentially leading to arbitrary code execution within the shell environment.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to read and process data from an Obsidian vault, which may contain untrusted content.
Ingestion points: Markdown file front matter located in /Users/mtm/Documents/Obsidian Vault (SKILL.md).
Boundary markers: Absent; there are no instructions provided to the agent to distinguish between its instructions and instructions found within the vault data.
Capability inventory: The skill possesses the ability to read files, execute shell commands, and modify file contents using the --mutate flag.
Sanitization: Absent; the skill does not suggest any validation or escaping of the data retrieved from the vault.
[DATA_EXPOSURE]: The skill relies on hardcoded absolute file paths targeting a specific user's directory (/Users/mtm/). While this is the stated purpose of the skill, it reveals the host's directory structure and could lead to unintended data access or functional issues if the environment does not match the hardcoded expectations.

islandiguana