The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes untrusted data from a local Obsidian vault.
Ingestion points: Processes note content, links, and folder structures within the target Obsidian vault as described in the execution flow section of SKILL.md.
Boundary markers: No specific boundary markers or instructions to ignore embedded commands within notes are present in the provided instructions.
Capability inventory: The skill aggregates data to produce a Markdown report and triggers external CLI queries via a dependency.
Sanitization: There is no evidence of sanitization for folder names or note metadata before they are used as parameters in queries.
[COMMAND_EXECUTION]: The skill's operational logic involves executing commands via a CLI tool.
Evidence: The "執行方式" (Execution method) section explicitly references a "CLI eval" pattern where query templates are populated with vault-specific parameters and executed locally.
Context: While intended for vault analysis, parameters derived from the vault (like folder paths) are passed into this evaluation process, which could be exploited if folder names contain shell-sensitive characters.

vault-report