菜单
Use when doing NixOS rebuilds involving taffybar, or when flake.lock updates are needed after changing taffybar ecosystem packages. Also use when debugging stale taffybar versions after `just switch`.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Safely reclaim disk space through reviewed, automated cleanup actions on this machine, especially for Nix, Rust build artifacts, containers, caches, and service state. Use when an assessment has identified cleanup candidates, the user asks to free space or automate remediation, or low-space recovery requires measured deletion of rebuildable data.
Measure and explain disk usage without deleting data, producing reusable compressed ncdu snapshots and an evidence-backed cleanup handoff. Use when disk is low, a filesystem is unexpectedly large, builds fail from space pressure, the user asks what is consuming space, or deeper iterative disk analysis is needed before cleanup.
Safely delegate between Codex and Claude through read-only CLI wrappers. Use when the user explicitly asks Codex or Claude to delegate to the other, or when a workflow explicitly calls for an independent cross-model second opinion.
Use when user wants to find promotional or unwanted recurring emails to unsubscribe from, or when doing periodic inbox hygiene to identify senders worth unsubscribing from
Use when the user wants to reset or rotate a website or service password end-to-end, including finding the right `pass` entry, generating a new password with `xkcdpassgen`, retrieving reset emails through `gws gmail` or a local mail CLI, completing the reset in the browser via the Chrome extension, and updating the password store safely without losing entry metadata.
Use when investigating production org-agenda-api state, testing endpoints, or debugging production issues
| name | taffybar-nixos-flake-chain |
| description | Use when doing NixOS rebuilds involving taffybar, or when flake.lock updates are needed after changing taffybar ecosystem packages. Also use when debugging stale taffybar versions after `just switch`. |
How the taffybar ecosystem packages are consumed by the NixOS configuration through a chain of nested flakes, and what flake.lock updates may be needed when something changes.
See also: taffybar-ecosystem-release for the package dependency graph, release workflow, and Hackage publishing.
The NixOS system build pulls in taffybar through the personal
imalison-taffybar config flake. The top-level NixOS flake should not declare
or override a direct taffybar input; the config flake owns its taffybar
version.
nixos/flake.nix (top - `just switch` reads this)
│ └── imalison-taffybar path:../dotfiles/config/taffybar
│
dotfiles/config/taffybar/flake.nix (middle - imalison-taffybar config)
│ ├── taffybar path:.../taffybar/taffybar
│ └── gtk-sni-tray, gtk-strut, etc. (GitHub inputs)
│
dotfiles/config/taffybar/taffybar/flake.nix (bottom - taffybar library)
│ └── gtk-sni-tray, gtk-strut, etc. (flake = false GitHub inputs)
The NixOS layer may make imalison-taffybar follow shared inputs such as
nixpkgs, flake-utils, and xmonad, but it should not set
imalison-taffybar.inputs.taffybar.follows.
path: inputs snapshot the target flake including its flake.lock at lock time. If you only run nix flake update at the top (nixos) layer, the middle and bottom layers keep whatever was previously locked in their own flake.lock files.
So when propagating a change to a system rebuild, you generally need to update flake.lock files from the bottom up — the bottom layer first so the middle layer picks up fresh locks when it re-resolves, then the middle so the top picks up fresh locks.
# Bottom (if an ecosystem dep changed):
cd ~/.config/taffybar/taffybar && nix flake update <pkg>
# Middle:
cd ~/.config/taffybar && nix flake update <pkg> taffybar
# Top:
cd /srv/dotfiles/nixos && nix flake update imalison-taffybar
Not every change requires touching all three layers. Think about which flake.lock files actually contain stale references:
nix flake update taffybar) then update imalison-taffybar at the top.kanshi-sni. Do not add a top-level taffybar input just to control the config flake's taffybar source.cd /srv/dotfiles/nixos && just switch
If taffybar seems stale after a rebuild, check whether the flake.lock at each layer actually points at the expected revision — a missed cascade step is the usual cause.