The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python utility (parser.py) via the Bash tool to analyze markdown files and generate callout blocks. This method provides deterministic results and avoids reliance on the LLM for complex parsing tasks.
[DATA_EXFILTRATION]: The analyze command in the provided script reads local markdown files to extract structured metadata (wikilinks and embeds). The extracted information is returned as JSON. The script does not access sensitive system files or perform network operations.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted markdown content from an Obsidian vault, creating a potential surface for indirect injection.
Ingestion points: Markdown files are read into the agent context via the obsidian-parser/parser.py script.
Boundary markers: None explicitly implemented; the parser relies on regex to target specific Obsidian syntax.
Capability inventory: The skill has permissions for Read, Write, and Bash execution.
Sanitization: The script uses strict regex patterns to extract only wikilinks and embeds, ensuring that surrounding prose or hidden instructions in the markdown do not interfere with the parsing logic.

obsidian-markdown-mastery