The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using nix develop or docker run to provide project-specific tools and runtimes. This is a core function but grants the skill significant control over the local execution environment.\n- [REMOTE_CODE_EXECUTION]: The skill fetches and executes remote Nix flakes from https://nixomatic.com. Since these flakes can contain arbitrary instructions for the Nix environment, this constitutes remote code execution via a vendor-managed service.\n- [EXTERNAL_DOWNLOADS]: The skill downloads environment configurations and Nix flakes from the external domain nixomatic.com during the execution of both Nix and Docker command templates.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its logic for managing development environments.\n
Ingestion points: The agent reads project manifest files (e.g., package.json, Cargo.toml) and the project README.md to identify required packages or existing environment URLs.\n
Boundary markers: There are no specific delimiters or instructions to ignore embedded commands within the files being analyzed.\n
Capability inventory: The skill possesses high-privilege capabilities including arbitrary shell command execution via nix and docker, as well as the ability to modify the project's README.md.\n
Sanitization: The skill lacks mechanisms to sanitize or validate the integrity of nixomatic.com URLs found in a project's documentation before reusing them, which could allow an attacker to trick the agent into running a malicious flake.

nixomatic