nix-profile-manager
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill provides instructions for the agent to install packages from remote Nix flakes, including arbitrary GitHub repositories (e.g., 'github:user/repo'). This results in the download and preparation of remote code for execution when the profile's binaries are used.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands for profile management, package discovery, and registry configuration, utilizing tools like the Nix CLI and jq.
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests and processes metadata from external flakes and registries. (1) Ingestion points: Output from 'nix search --json' and 'nix registry list'. (2) Boundary markers: Absent in the provided instructions for data parsing. (3) Capability inventory: File system access, binary installation via 'nix profile add', and path modification. (4) Sanitization: The instructions do not include methods for sanitizing package metadata from potentially untrusted remote sources.
Audit Metadata