The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted Markdown files from an external source ('inbox' clippings).
Ingestion points: Reads Markdown files directly from ~/Code/personal/obsidian/llm-wiki/inbox/.
Boundary markers: Absent. The workflow reads the content of these notes without using delimiters or instructions to ignore embedded commands.
Capability inventory: The agent has the ability to read, write, and edit files, as well as move files and execute shell commands like find, mv, and qmd.
Sanitization: Absent. There is no evidence of filtering or validation of the text extracted from the inbox notes before it is used to generate new wiki content.
[COMMAND_EXECUTION]: The skill utilizes several shell commands to manage the vault environment. It uses find to locate notes, mv to move processed files, and a non-standard command qmd update to re-index the vault. These commands are scoped to the specific vault path.

obsidian-process-notes