The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes user-provided transcription data to generate metadata and observations.
Ingestion points: Transcription content provided by the user in instructions/transcription.md.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the transcription text when generating metadata.
Capability inventory: The skill utilizes Obsidian:write_note, Obsidian:read_note, Obsidian:patch_note, and Obsidian:search_notes tools.
Sanitization: The instructions explicitly require the transcription body to be preserved verbatim without modification or sanitization.
[COMMAND_EXECUTION]: The skill performs shell-based configuration and environment setup during its initial execution.
Evidence: The agent is instructed to use mkdir, ln -s, and printf to create a global configuration at ~/.config/wrap-up/vault and local project symlinks in references/bootstrap.md.
Evidence: The skill uses git rev-parse --show-toplevel to resolve the current project root in references/mapping.md.
Purpose: These operations are limited to establishing and maintaining the project-to-vault registry system.

notes