The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of Nushell scripts and mise tasks to automate the discovery, triage, and update process for skill dependencies.
[EXTERNAL_DOWNLOADS]: Queries well-known services and package registries, including GitHub (api.github.com), Hex.pm, and Crates.io, to retrieve current version numbers and release metadata. These operations are conducted using built-in Nushell HTTP commands.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it parses sources.toml files which may contain untrusted data from various plugins.
Ingestion points: Plugin configuration files (sources.toml) and project metadata (marketplace.json).
Boundary markers: Absent; configuration values such as URLs and version strings are interpolated directly into script logic.
Capability inventory: Network request capabilities (http get, http head) and file system write operations (save --force).
Sanitization: Absent; URLs retrieved from configuration files are utilized in network requests without formal validation or escaping.

skill-update