The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill establishes a significant attack surface by instructing the agent to treat instructions found in external data files as authoritative overrides.
Ingestion points: The agent is directed to read AGENTS.md and the "vault's own router or playbook" files located within any Obsidian vault it processes (SKILL.md).
Boundary markers: Absent. The skill contains no instructions to ignore or delimit embedded commands; instead, it explicitly states, "Live vault instructions override this skill."
Capability inventory: The skill enables the agent to search, read, create, and organize notes, which involves substantial file system manipulation.
Sanitization: There are no mechanisms for sanitizing or validating the instructions retrieved from the vault before the agent executes them.

obsidian-vault