nix-workflow

Pass

Audited by Gen Agent Trust Hub on Oct 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructions involve reading external content that could be manipulated by a third party to influence the agent's behavior.
  • Ingestion points: The skill recommends reading repository-specific instructions (SKILL.md), searching external GitHub issues and pull requests (references/troubleshooting.md), and analyzing build logs (references/troubleshooting.md).
  • Boundary markers: The instructions lack specific delimiters or warnings to the agent to ignore potentially malicious instructions embedded in the external content retrieved from searches or logs.
  • Capability inventory: The skill utilizes powerful tools such as nix, nix develop, nix build, and gh which provide deep system and network access.
  • Sanitization: There are no explicit sanitization or filtering steps described for processing external content before the agent uses it to propose fixes or configuration changes.
  • [EXTERNAL_DOWNLOADS]: The skill references a wide range of external tools, frameworks, and documentation sources within the Nix ecosystem.
  • Evidence: Includes links to various repositories under the github.com/nix-community, github.com/NixOS, github.com/numtide, and github.com/DeterminateSystems organizations.
  • Description: These references target well-known services and organizations that provide standard utilities for Nix development and are documented neutrally as resources for the user.
  • [COMMAND_EXECUTION]: The primary functionality of the skill involves providing instructions for executing shell commands to interact with the Nix package manager and system configuration.
  • Evidence: The reference files frequently use commands such as nix, nix-store, nix-locate, and gh for querying the environment, managing packages, and debugging.
  • Description: While the skill requires significant shell interaction, it includes safety instructions directing the agent to seek explicit user authorization before performing operations that change system state or modify sensitive configurations like nix.conf.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 10, 2026, 03:53 PM