The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill includes a hardcoded API key (d4fc944e64f535df05b34c2e8596c4e3eaffc6dfd5b5ee15e76789a4ab96698a) for the Obsidian Local REST API. Hardcoding credentials in skill files is an insecure practice as it exposes secrets to the agent's context.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from an external source (the Obsidian vault) which may contain malicious instructions. 1. Ingestion points: Reads markdown files, index files, and workflow guidelines from the vault directory at /Users/espen/Library/CloudStorage/OneDrive-Espennilsen.net/2-Areas/Digital_Life/Obsidian/e9n. 2. Boundary markers: The skill does not implement boundary markers or instructions to ignore embedded commands within the vault files. 3. Capability inventory: The agent has the ability to read, create, update, and delete files via the REST API, and use shell commands for filesystem search and retrieval. 4. Sanitization: No sanitization or validation of the content retrieved from the vault is specified.
[COMMAND_EXECUTION]: The skill providing templates for executing shell commands including curl, cat, find, and grep to interact with the local API and the filesystem to manage vault data.

obsidian-vault