The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it reads and summarizes external note data.
Ingestion points: Data is ingested via obsidian_get_file_contents and obsidian_batch_get_file_contents tools.
Boundary markers: The skill does not implement delimiters or 'ignore' instructions when processing vault content.
Capability inventory: The skill has permissions to write to the file system using obsidian_append_content and obsidian_patch_content.
Sanitization: Retrieved content is not sanitized or escaped before being processed by the agent.
[EXTERNAL_DOWNLOADS]: The configuration guide directs users to download a package from the npm registry.
Evidence: Setup instructions recommend running npx -y mcp-obsidian.
Context: This targets a well-known service (npm) for standard tool installation.
[COMMAND_EXECUTION]: The documentation provides bash commands for the user to configure their local environment.
Evidence: Commands such as mkdir -p ~/.claude and cat >> ~/.claude/mcp.json are presented in the setup section.

obsidian-helper