The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script named claude-obsidian.py using python3 to audit an Obsidian vault. The script is located within a predefined product directory (/absolute/path/to/installed/claude-obsidian). This is a standard implementation of a local utility tool.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from an external Obsidian vault, which could contain untrusted content.
Ingestion points: The contents of the Obsidian vault files located at the path provided in the $VAULT environment variable.
Boundary markers: The instructions explicitly direct the agent to "Observe the deterministic report" and "Preserve the engine's paths, line numbers, targets, categories, and counts," effectively anchoring the agent's behavior to the structured output of the tool.
Capability inventory: The skill uses python3 to execute a local script and requires read access to the filesystem to scan the vault.
Sanitization: The skill relies on the deterministic nature of the claude-obsidian.py parser to identify specific structural issues (dead links, orphan pages) rather than performing semantic analysis on the text, which limits the influence of any instructions embedded within the vault data.

wiki-lint