The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through its processing of untrusted Obsidian vault files.
Ingestion points: The read_canvas function in scripts/canvas_ops.py, the read_note function in scripts/vault_ops.py, and the analyze command in obsidian-parser/parser.py all ingest external file content.
Boundary markers: Script outputs do not include boundary markers or explicit instructions to distinguish vault content from system instructions.
Capability inventory: The skill has the ability to perform file writes and renames via scripts/canvas_ops.py and scripts/vault_ops.py.
Sanitization: Content retrieved from vault files is returned to the agent without filtering or sanitization.
[COMMAND_EXECUTION]: The skill performs file modifications including writing and renaming using Python's os and pathlib modules. While these are necessary for its primary purpose, the implementation lacks path validation for the --file input, potentially allowing an agent to manipulate files at any path accessible to the script environment.

obsidian-canvas-architect