The Agent Skills Directory

[DATA_EXFILTRATION]: The skill contains a hardcoded absolute file path which exposes sensitive system metadata including the local username and directory layout.
Evidence: The path /Users/mtm/Documents/Obsidian Vault/CLAUDE.md provides information about the host environment's user identity ('mtm') and file organization (SKILL.md).
[INDIRECT_PROMPT_INJECTION]: The skill's workflow involves reading and modifying external data from Obsidian notes, which introduces a risk of the agent following malicious instructions embedded in those notes.
Ingestion points: The agent is instructed to find and read notes carrying the storereview tag within the Obsidian vault (SKILL.md).
Boundary markers: Absent; there are no clear delimiters or instructions to the model to ignore instructions discovered within the notes' contents.
Capability inventory: The skill utilizes file-system affecting tools including git commit and obsidian-cli (SKILL.md).
Sanitization: Absent; the skill does not include steps to sanitize or validate the content of the notes before applying changes such as heading renaming or committing the file to version control.

store-review-fix