The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill clones a project repository from the author's GitHub account (cylqwe7855-alt/personal-wiki) to the local environment during the initialization phase.
[REMOTE_CODE_EXECUTION]: Following the repository clone, the skill automatically installs Python and Node.js dependencies from the downloaded files using 'pip install' and 'npm install', which executes third-party code from public registries.
[COMMAND_EXECUTION]: The skill executes various shell commands to manage virtual environments, run ingestion scripts, and launch a local web server via 'npm run dev'. Notably, for 'Other' data sources, it dynamically writes and executes custom Python scripts.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) because it processes untrusted data from user notes and documents to synthesize wiki content.
Ingestion points: Data enters the system from Obsidian vaults, Apple Notes databases, PDF/Docx files, and the 'raw/' directory.
Boundary markers: There are no specified delimiters or 'ignore' instructions to prevent malicious commands embedded in notes from influencing the agent's writing behavior.
Capability inventory: The skill possesses extensive capabilities including file system modification, command execution (git, pip, npm, python), and local network server operations.
Sanitization: The skill does not appear to sanitize or validate external content before it is read and used to update the wiki articles or generate ingestion logic.

wiki