refactor(nixos): scope agent guidance

.pi/prompts/tldr.md

@@ -0,0 +1,8 @@
+---
+description: Summarize an agent response
+argument-hint: "[source request]"
+---
+
+Summarize `${ARGUMENTS:-the preceding agent response}`.
+Keep the summary compact.
+Use no fixed format.

.pi/skills/agent-writing-guide

@@ -0,0 +1 @@
+../../src/skills/agent-writing-guide

.pi/skills/craft-skill

@@ -1 +0,0 @@
-../../src/skills/craft-skill

.pi/skills/retrospect

@@ -0,0 +1 @@
+../../src/skills/retrospect

.pi/skills/to-document

@@ -1 +0,0 @@
-../../src/skills/to-document

AGENTS.md

@@ -48,6 +48,8 @@
 - NEVER ask the user a question using the `AskUserQuestion` tool.
   Ask in plain prose, in your own message, instead, with no exceptions.
   This overrides any default instruction to use it.
+- For review requests, inspect the requested material and report findings.
+  Edit files after the operator accepts a proposed change.
 - Never manually modify `CHANGELOG.md` files or files marked as auto-generated.
   Detect auto-generated files with a layered check.
   Trust an explicit in-file marker first, for example `AUTO-GENERATED, DO NOT EDIT`.
@@ -67,20 +69,6 @@
   after applying an accepted change.
   If the workflow is paused, say that plainly instead of using a dangling transition like "continuing" or "next".
 
-## Corrections
-
-- When you discover that your belief about a project fact or convention was wrong, record the correction.
-  Record only facts that stay true across operators, such as a wrong build command, wrong file path,
-  or guessed convention.
-  Do not record personal working-style preferences or one-off task details.
-- Record the correction in the nearest project `AGENTS.md`.
-  Use a dedicated `## Gotchas` section.
-  Create that section when the file lacks one.
-- If no project `AGENTS.md` exists, ask before creating one.
-  Append the correction immediately without asking when a project `AGENTS.md` exists.
-  Briefly mention the edit in your response.
-  If an existing entry is wrong or stale, correct or remove it the same way.
-
 ## Gotchas
 
 - When materializing an external specification, standard, or project-maintained convention, use the
@@ -100,8 +88,6 @@
 - Ariadne Rust tests are declared in Bazel `BUILD` files; do not add `[[test]]` entries to
   `src/ariadne/Cargo.toml` unless explicitly requested.
 - In this monorepo, "dotfiles" refers to the dotfiles/NixOS configuration under `src/dotfiles/` and `configs/nixos/`.
-- The flake does not currently define `formatter.x86_64-linux`.
-  Do not use `nix fmt` as a validation command until a formatter is added.
 - The `fix` app passes its current directory to Deadnix.
   Run `nix run .#fix` from the repository root so path exclusions match.
 - Git pathspec searches started in a subdirectory default to that subtree.
@@ -124,3 +110,7 @@
   Preserve `git-subtree-dir` and `git-subtree-split` trailers on Pi vendor commits.
 - The default Pi model for NixOS user configuration is set in `configs/nixos/modules/shell/pi/default.nix`, not in vendored Pi source.
 - The `monorepo` and `status` libraries do not have `README.md` files.
+- Check root and `src/skills/.pi/skills/` for enabled symlinks before removing a monorepo skill.
+- Ariadne Rust source files live directly under `src/ariadne/`. The binary entry point is `ariadne_main.rs.
+- Bazel build-style rules live under `src/skills/style/build/rules/`, not `src/skills/style/rules/`.
+- Bazel passes arguments after `--` to the executed binary verbatim. A scaffold can define `//path` as a destination-path syntax.

configs/nixos/AGENTS.md

@@ -1,70 +0,0 @@
-# NixOS configuration instructions
-
-## Desktop
-
-- When adding a graphical application, register its workspace icon in `registry.window-icon`.
-  The workspace indicator uses this mapping to show an application icon instead of the generic default glyph.
-  Match the mapping on the window class.
-  Store the icon as the rendered string.
-  Run `hyprctl clients -j | jq -r '.[].class' | sort -u` to list classes in the running session.
-- TODO: Verify this path after the NixOS configuration is migrated.
-  Set Hyprland session environment variables in `modules/desktop/session/hyprland/default.nix` with `env = KEY,VALUE`.
-  `home.sessionVariables` do **not** reach the Hyprland session because UWSM does not source `hm-session-vars.sh`.
-  Source cursor values from `config.stylix.cursor`.
-  Set `XCURSOR_*` and `HYPRCURSOR_*` to the same theme for Bibata.
-
-## Firefox
-
-- Manage Firefox search engines through `programs.firefox.search` with `force = true`.
-  Hide built-in engines with `<engine>.metaData.hidden = true` because omission does not remove them.
-  Use current engine ids, such as `default = "ddg"` instead of `"DuckDuckGo"`.
-  Decode the built `search.json.mozlz4` with `mozlz4a -d <search.json.mozlz4>` to check the result.
-- Avoid Home Manager Firefox extension settings unless the whole profile can use the legacy extension-storage backend.
-  Any non-empty `profiles.<name>.extensions.settings.<id>.settings` sets `extensions.webextensions.ExtensionStorageIDB.enabled = false` for the whole profile.
-
-## Neovim
-
-- Nixvim's flake input follows the root nixpkgs source, but its Home Manager module does not reuse the host's `pkgs` instance.
-  Keep `programs.nixvim.nixpkgs.useGlobalPackages = true`.
-  That setting makes Nixvim use the shared package set without warning that its source default changed.
-
-## AI tooling
-
-- Use produced literals for Herdr key names that include shifted punctuation.
-  For example, use `prefix+<` instead of `prefix+shift+comma`.
-- Put monorepo Pi extension sources in `.pi/extensions/`.
-  Pi discovers them from subdirectories through the local ancestor-resource patch.
-- Treat Home Manager-managed Pi settings as generated output.
-  Do not edit them except for throwaway development or local experiments.
-  Promote finished settings into the NixOS configuration before treating them as deployed.
-
-## Gotchas
-
-- `configs/nixos/registry/stable-uid.nix` receives library functions through `lib`.
-  Qualify its library calls with `lib.`.
-- Guest modules are host-agnostic wiring, while host files own deployment-specific service configuration such as served volumes and host paths.
-  For example, `guests/copyparty.nix` should expose options and translate them into guest mounts/service settings, while `hosts/pikachu/default.nix` chooses `/inbox`, `/media`, and their backing paths.
-  Guest addresses and routes are host-level configuration because a guest module can be used on multiple hosts.
-- `sops.secrets.<name>.owner` must name a user known on the decrypting host.
-  Do not set the owner to a guest-only service user unless the host also declares that user.
-  Use a host-visible owner plus a shared group mode when the guest service can read the decrypted secret file through its group.
-- A read-only guest bind mount cannot contain nested writable bind mounts unless their mount-point
-  directories already exist inside the read-only source.
-  For aggregate Copyparty volumes such as `/srv` with nested `/srv/inbox`, keep the guest bind mount
-  writable and enforce read-only behavior through Copyparty access plus service `ReadWritePaths`.
-- `deploy-to` builds from the source checkout, records the imported closure as the target system profile,
-  and activates it. Do not rely on a target-local monorepo checkout for remote deployment.
-- On Pikachu, `nix copy --to ssh://pikachu` accepted a small `nix store add-file` probe,
-  but rejected locally built system paths that lacked trusted signatures.
-  Use a sudo `nix-store --import` path for unsigned deploy closures.
-- Public DNS entries are not managed through the NixOS configuration.
-  Do not describe NixOS route choices as public-DNS management.
-- Do not hand-author `hardware-configuration.nix` for a host that has not been generated on that machine.
-  Generate it on the target system or live ISO, then stage it before evaluating the flake.
-- Configure the Chaotic Nyx cache only on hosts that select `linuxPackages_cachyos` (currently Neogaia
-  and Zeus), not in the generic host module.
-- Module-relative paths resolve from the module file. Update them when moving a module into a subdirectory.
-- On Pikachu, `uptime` does not support `-s`. Use `who -b` to report the boot time.
-- Evaluating a NixOS guest's inner `config.services.nginx` can force incomplete ACME values.
-  Evaluate the parent host's `config.system.build.toplevel` to validate an ACME-enabled guest.
-- `pkgs.systemd` does not set `meta.mainProgram`. Use `lib.getExe' pkgs.systemd "systemctl"` instead of `lib.getExe pkgs.systemd`.

configs/nixos/guests/reverse-proxy.nix

@@ -13,11 +13,6 @@ in
         modules.services.reverse-proxy.enable = true;
 
         services.nginx.virtualHosts = {
-          "games.alexion.dev" = {
-            forceSSL = true;
-            enableACME = true;
-            locations."/".proxyPass = "http://10.23.20.32";
-          };
           "git.alexion.dev" = {
             forceSSL = true;
             enableACME = true;

configs/nixos/guests/romm.nix

@@ -0,0 +1,181 @@
+{
+  config,
+  lib,
+  ...
+}:
+let
+  cfg = config.guests.romm;
+  rommUid = config.registry."stable-uid".${config.networking.hostName}.romm;
+  redisUid = config.registry."stable-uid".${config.networking.hostName}."redis-romm";
+in
+{
+  options.guests.romm = {
+    dataPath = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/srv/romm";
+      description = "Host path mounted as RomM application data.";
+    };
+
+    postgresqlPath = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/srv/romm/postgresql";
+      description = "Host path mounted as RomM's PostgreSQL data.";
+    };
+
+    redisPath = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/srv/romm/redis";
+      description = "Host path mounted as RomM's Redis data.";
+    };
+
+    romsPath = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/srv/roms";
+      description = "Host path mounted as RomM's ROM library.";
+    };
+
+    biosPath = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/srv/bios";
+      description = "Host path mounted as RomM's BIOS library.";
+    };
+
+    usersPath = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/srv/romm/users";
+      description = "Host path mounted as RomM's user assets.";
+    };
+
+  };
+
+  config = lib.mkIf cfg.enable {
+    assertions =
+      map
+        (path: {
+          assertion = path.value != null;
+          message = "guests.romm.${path.name} must point at the host directory that stores ${path.description}.";
+        })
+        [
+          {
+            name = "dataPath";
+            value = cfg.dataPath;
+            description = "RomM application data";
+          }
+          {
+            name = "postgresqlPath";
+            value = cfg.postgresqlPath;
+            description = "RomM PostgreSQL data";
+          }
+          {
+            name = "redisPath";
+            value = cfg.redisPath;
+            description = "RomM Redis data";
+          }
+          {
+            name = "romsPath";
+            value = cfg.romsPath;
+            description = "RomM's ROM library";
+          }
+          {
+            name = "biosPath";
+            value = cfg.biosPath;
+            description = "RomM's BIOS library";
+          }
+          {
+            name = "usersPath";
+            value = cfg.usersPath;
+            description = "RomM's user assets";
+          }
+        ];
+
+    guests.romm = {
+      interior = {
+        modules.services.romm.enable = true;
+        services.romm.nginx.virtualHost = cfg.reverseProxy.route.domain;
+      };
+
+      mounts = {
+        "/var/lib/romm" = {
+          directory = cfg.dataPath;
+          owner = rommUid;
+          group = "storage";
+          mode = "2770";
+        };
+        "/var/lib/postgresql" = {
+          directory = cfg.postgresqlPath;
+          owner = config.ids.uids.postgres;
+          group = "storage";
+          mode = "0700";
+        };
+        "/var/lib/redis-romm" = {
+          directory = cfg.redisPath;
+          owner = redisUid;
+          group = "storage";
+          mode = "0700";
+        };
+        "/var/lib/romm/library/roms" = {
+          directory = cfg.romsPath;
+          owner = rommUid;
+          group = "storage";
+          mode = "2770";
+        };
+        "/var/lib/romm/library/bios" = {
+          directory = cfg.biosPath;
+          owner = rommUid;
+          group = "storage";
+          mode = "2770";
+        };
+        "/var/lib/romm/assets/users" = {
+          directory = cfg.usersPath;
+          owner = rommUid;
+          group = "storage";
+          mode = "2770";
+        };
+        "/run/secrets/romm.env" = {
+          file = config.sops.templates."romm.env".path;
+          readOnly = true;
+        };
+      };
+
+      limits = {
+        memory = lib.mkDefault "4G";
+        cpu = lib.mkDefault "200%";
+        tasksMax = lib.mkDefault 512;
+      };
+    };
+
+    sops.secrets =
+      lib.genAttrs
+        [
+          "services/romm/igdb/client-id"
+          "services/romm/igdb/client-secret"
+          "services/romm/retroachievements/api-key"
+          "services/romm/screenscraper/user"
+          "services/romm/screenscraper/password"
+          "services/romm/steamgriddb/api-key"
+        ]
+        (_: {
+          sopsFile = ../secrets + "/${config.networking.hostName}.yaml";
+        });
+
+    sops.templates."romm.env" = {
+      owner = "root";
+      group = "storage";
+      mode = "0440";
+      content = ''
+        IGDB_CLIENT_ID=${config.sops.placeholder."services/romm/igdb/client-id"}
+        IGDB_CLIENT_SECRET=${config.sops.placeholder."services/romm/igdb/client-secret"}
+        SCREENSCRAPER_USER=${config.sops.placeholder."services/romm/screenscraper/user"}
+        SCREENSCRAPER_PASSWORD=${config.sops.placeholder."services/romm/screenscraper/password"}
+        RETROACHIEVEMENTS_API_KEY=${config.sops.placeholder."services/romm/retroachievements/api-key"}
+        STEAMGRIDDB_API_KEY=${config.sops.placeholder."services/romm/steamgriddb/api-key"}
+      '';
+    };
+  };
+}

configs/nixos/hosts/neogaia/default.nix

@@ -15,6 +15,8 @@
 
   modules.host.cachyos.enable = true;
 
+  services.xserver.xkb.options = "caps:swapescape";
+
   modules.network.enable = true;
 
   registry.address.neogaia.self = "10.23.50.146";

configs/nixos/hosts/pikachu/default.nix

@@ -88,6 +88,26 @@
     };
   };
 
+  guests.romm = {
+    enable = true;
+    vlan = 20;
+    address = "10.23.20.112";
+    dataPath = "/pikachu/data/romm";
+    postgresqlPath = "/pikachu/data/romm/postgresql";
+    redisPath = "/pikachu/data/romm/redis";
+    romsPath = "/pikachu/media/games/emulation/roms";
+    biosPath = "/pikachu/media/games/emulation/bios";
+    usersPath = "/pikachu/media/games/emulation/users";
+    reverseProxy = {
+      enable = true;
+      route = {
+        domain = "games.alexion.dev";
+        protocol = "https";
+        port = 80;
+      };
+    };
+  };
+
   guests.reverse-proxy = {
     enable = true;
     vlan = 20;

configs/nixos/modules/desktop/apps/firefox.nix

@@ -67,6 +67,8 @@ in
 
           # Stylix uses a near-white selection background.
           # Use the Nord selection color because its text stays readable.
+          # Home Manager writes extension settings through Firefox's legacy storage backend.
+          # Extension settings disable `ExtensionStorageIDB` for the whole profile.
           extensions.settings."FirefoxColor@mozilla.com".settings.theme.colors.popup_highlight =
             let
               c = hm.config.lib.stylix.colors;

configs/nixos/modules/desktop/session/hyprland/default.nix

@@ -37,6 +37,7 @@
         config = {
           input = {
             kb_layout = config.services.xserver.xkb.layout;
+            kb_options = config.services.xserver.xkb.options;
             repeat_delay = 250;
             repeat_rate = 45;
             accel_profile = "flat";

configs/nixos/modules/desktop/session/rofi.nix

@@ -28,12 +28,11 @@ in
       # Stylix supplies the Nord theme.
       programs.rofi = {
         enable = true;
-        font = lib.mkForce "JetBrains Mono 12";
-
         # Application launch is the only mode, kept minimal so the prompt opens
         # as fast as rofi can read the desktop entries.
         # Icons are off: resolving one per entry is the bulk of drun's startup.
-        extraConfig = {
+        settings = {
+          font = lib.mkForce "JetBrains Mono 12";
           modi = "drun";
           show-icons = false;
         };

configs/nixos/modules/services/romm.nix

@@ -0,0 +1,47 @@
+{
+  config,
+  lib,
+  ...
+}:
+let
+  cfg = config.modules.services.romm;
+in
+{
+  options.modules.services.romm = {
+    enable = lib.mkEnableOption "RomM";
+
+    environmentFile = lib.mkOption {
+      type = lib.types.path;
+      default = "/run/secrets/romm.env";
+      readOnly = true;
+      description = "Path to RomM's environment file.";
+    };
+  };
+
+  config = lib.mkMerge [
+    {
+      registry."stable-uid".${config.networking.hostName} = {
+        romm = 20005;
+        "redis-romm" = 20006;
+      };
+    }
+
+    (lib.mkIf cfg.enable {
+      networking.firewall.allowedTCPPorts = [ 80 ];
+
+      services.romm = {
+        enable = true;
+        group = "storage";
+        database.createLocally = true;
+        redis.createLocally = true;
+        watcher.enable = true;
+        inherit (cfg) environmentFile;
+      };
+
+      users.users = {
+        romm.uid = config.registry."stable-uid".${config.networking.hostName}.romm;
+        redis-romm.uid = config.registry."stable-uid".${config.networking.hostName}."redis-romm";
+      };
+    })
+  ];
+}

configs/nixos/modules/shell/nvim/default.nix

@@ -24,6 +24,9 @@ in
     home-manager.users.${user} = hm: {
       programs.nixvim = {
         enable = true;
+
+        # Aligns Nixvim plugins with this flake's `Nixpkgs`.
+        # `nixpkgs.useGlobalPackages` prevents Home Manager from creating a separate package set.
         nixpkgs.useGlobalPackages = true;
 
         extraPackages = with pkgs; [

configs/nixos/registry/stable-uid.nix

@@ -12,7 +12,7 @@ let
 
 in
 {
-  # Next stable UID: 20005
+  # Next stable UID: 20007
 
   type = lib.types.ints.positive;
 

configs/nixos/secrets/pikachu.yaml

@@ -6,6 +6,17 @@ services:
         lastfm:
             api-key: ENC[AES256_GCM,data:Zb0I5TpdEqOxLAD6ed2XywFGZp7dfY69vcGiWOUJkJQ=,iv:pO53pHr09jWfzJyTsQbTHtzTmiQ9+RQkDwZJmInIiX8=,tag:io9jkSN7TF2I8+qlSymaYw==,type:str]
             secret: ENC[AES256_GCM,data:rKhd9i0/tbJlXhc9Wf81IDydjpD8khSdwJEEMK260E4=,iv:btBXWBKJp9br5a3yZsRVegN8Ytzc45Qbk2iNlPbENkQ=,tag:LUbu+6gxfqQvVpddZto7Tg==,type:str]
+    romm:
+        igdb:
+            client-id: ENC[AES256_GCM,data:tme7NzzbOlCmkMwVObWz793mOsCyVIzSjGoFvxpe,iv:5aOMvsDi8Z9aBymbqhbK79X4kJsqaET1y2Py6YUZZ/U=,tag:ih3GghlxiLsVyeTJoqU9/Q==,type:str]
+            client-secret: ENC[AES256_GCM,data:AqF7B9VaxM/X3sRqY6C5JVWPWqgRIoqy09P9ZUFj,iv:D+vDPUuGGdzKYnTVTRkI9TV/EOJgVECIeNDt9g8NGl0=,tag:MRaTT7TdEZSd+98xthVw1A==,type:str]
+        retroachievements:
+            api-key: ENC[AES256_GCM,data:LMzETQCMi9Ib783W0BZDe2PaSUSTVM5a61502RtZLg8=,iv:Hf5nBoVtZdJszQuc2n0LVNMtJCCpJrh40y4g+l+az70=,tag:esfKLpI1W8BG3Pfsg3xxKQ==,type:str]
+        steamgriddb:
+            api-key: ENC[AES256_GCM,data:+kBRieIB/dHbvQLbH2gQTQh2j6DQhybVWEGkSCRZrZY=,iv:3h6dhfiV7HfaZhMv93x8+7F3utzYG6fIbTDScA14PVI=,tag:Wo24YnvsIxe7iMRLyxrj2w==,type:str]
+        screenscraper:
+            user: ENC[AES256_GCM,data:7hGcFF4xqg==,iv:SFOdYej94Thfn+Te0kXsC9Kq568A1dsnaR+JB90MgBQ=,tag:inIWuFY7f5Tvv8EyV71dYA==,type:str]
+            password: ENC[AES256_GCM,data:eUWwDDQUSytm/6QZlG5XjbSi41nsBQbm4w==,iv:82S2a3rAVjaF0CXcjXUXErNHMY7MEkjpDL7EVX9cIHw=,tag:uGuVJCv8ThsaNnA0C0WR2Q==,type:str]
 sops:
     age:
         - enc: |
@@ -26,7 +37,7 @@ sops:
             dyjzy0/9Ui1IIXH36AImO5hiEZcrGePV2qPLLQc7ZLngC1jlJuEKJA==
             -----END AGE ENCRYPTED FILE-----
           recipient: age1wf5s0n0tgt6ld2ysgu9dc67mj8ylwecgl4utzg7hqwy3kut9zyms7aglmh
-    lastmodified: "2026-09-12T22:54:25Z"
+    lastmodified: "2026-10-20T01:27:06Z"
-    mac: ENC[AES256_GCM,data:4R3a0ytf4i8aM8iBQ9vRC/q+hvtjLrCTNgbpCB5DbqBMBYmhtGsBmO/0i1yT0N4tydyLaGOMkNfuuuVi3ASLEOKnViy9w3uD9IN0F79W4f2vguWUVIp/T8skvN9wbV9o9xd0LkJzmeVVgzb1H6HKw+rTGjbquleoIHEBDcDDM+k=,iv:0NlckMXlS+idY6UGBF+F0KFLt+pzD7Q/YEQOdXraAYg=,tag:JBBQduV8uhnKz9gPf6HfAQ==,type:str]
+    mac: ENC[AES256_GCM,data:NCt6e/ctZwaBADfrhu1RXtILlClbHEqOAnIu8rE/zMPSwMIPB/kTtfd/ujf+1iRszoHaX4Se1xc0wz1th6NVToFVNarxBBgTk4JMUU8gdMweOjmwlh3GXU3ZcYY/xAXxWF1hxyXSEW78j3S2844H9XXENAq/Bf3mjdCYQ/qr/d4=,iv:4nFyBSmRvF4zSAwvL+xccMwtWrVbHtFKNoTb7f3rSH4=,tag:eCV/7P0TjyMZXKcWBUPXQA==,type:str]
     unencrypted_suffix: _unencrypted
     version: 3.13.3

docs/.obsidian/graph.json

@@ -74,6 +74,6 @@
   "repelStrength": 20,
   "linkStrength": 1,
   "linkDistance": 250,
-  "scale": 0.24304046676817048,
+  "scale": 0.16202697784544698,
   "close": false
 }

docs/.obsidian/workspace.json

@@ -8,17 +8,17 @@
         "type": "tabs",
         "children": [
           {
-            "id": "94a7b5a014888797",
+            "id": "daab69fc3faab88f",
             "type": "leaf",
             "state": {
               "type": "markdown",
               "state": {
-                "file": "03 wrk/tasks/0002-nixos-refactor.md",
+                "file": "00 INBOX/credits.md",
                 "mode": "source",
                 "source": true
               },
               "icon": "lucide-file",
-              "title": "0002-nixos-refactor"
+              "title": "credits"
             }
           }
         ]
@@ -91,7 +91,7 @@
             "state": {
               "type": "localgraph",
               "state": {
-                "file": "03 wrk/tasks/0002-nixos-refactor.md",
+                "file": "00 INBOX/credits.md",
                 "options": {
                   "collapse-filter": true,
                   "search": "",
@@ -217,34 +217,34 @@
       "table-editor-obsidian:Advanced Tables Toolbar": false
     }
   },
-  "active": "94a7b5a014888797",
+  "active": "daab69fc3faab88f",
   "lastOpenFiles": [
-    "TODO.md",
+    "00 INBOX/ai/tldr-design.md",
-    "03 wrk/productivity.md",
+    "00 INBOX/ai/caps-escape-design.md",
-    "03 wrk/tasks/0003-whisper-dictation.md",
+    "AGENTS.md",
-    "03 wrk/tasks/0001-todo-cli.md",
+    "00 INBOX/ai/tldr-design-tree.md",
+    "00 INBOX/ai/grill-skill-design.md",
+    "03 wrk/tasks/0012-gitea-axi-pr-merge-guardrail.md",
+    "03 wrk/tasks/0011-gitea-axi-rust-migration.md",
+    "00 INBOX/ai/gitea-axi-pr-merge-guardrail.md",
+    "00 INBOX/ai/gitea-axi-rust-migration.md",
+    "00 INBOX/valheim progression.md",
+    "03 wrk/tasks/0010-fix-main-monitor-detection.md",
+    "00 INBOX/credits.md",
+    "03 wrk/tasks/0009-bazel-project-scaffolding.md",
+    "03 wrk/tasks/0005-rust-cli-scaffolding.md",
+    "03 wrk/tasks/0008-vim-obsidian-markdown-renderer.md",
     "03 wrk/tasks/0002-nixos-refactor.md",
-    "03 wrk/dotfiles.md",
+    "TODO.md",
-    "03 wrk/machoke.md",
+    "technical-writing.md",
-    "03 wrk/falinks.md",
+    "dictionary.md",
-    "03 wrk/ariadne.md",
+    "03 wrk/tasks/0007-obsidian-source-notes.md",
-    "99 meta/templates/task.md",
+    "03 wrk/tasks/0004-dictionary-cli.md",
-    "03 wrk/tasks/0005-modules-services-navidrome.md",
+    "03 wrk/tasks/0006-wallpaper-randomizer.md",
-    "03 wrk/tasks/0003-registry-homepage.md",
+    "03 wrk/tasks/0003-whisper-dictation.md",
-    "03 wrk/tasks/0039-hosts-zeus-hardware-configuration.md",
+    "03 wrk/monorepo.md",
-    "03 wrk/tasks/0030-dotfiles-hosts.md",
+    "03 wrk/tasks/0001-todo-cli.md",
-    "03 wrk/tasks/0038-hosts-zeus-disk.md",
+    "03 wrk/productivity.md",
-    "03 wrk/tasks/0037-hosts-zeus-default.md",
-    "03 wrk/tasks/0036-hosts-pikachu-hardware-configuration.md",
-    "03 wrk/tasks/0035-hosts-pikachu-disk.md",
-    "03 wrk/tasks/0034-hosts-pikachu-default.md",
-    "03 wrk/tasks/0033-hosts-neogaia-hardware-configuration.md",
-    "03 wrk/tasks/0032-hosts-neogaia-disk.md",
-    "03 wrk/tasks/0031-hosts-neogaia-default.md",
-    "03 wrk/tasks/0029-guests-valheim.md",
-    "03 wrk/tasks/0028-guests-reverse-proxy.md",
-    "03 wrk/tasks/0027-guests-palworld.md",
-    "03 wrk/tasks/0026-guests-navidrome.md",
     "03 wrk/tasks",
     "00 INBOX/let order name resolution.md~",
     "00 INBOX/nix name resolution.md~",

docs/00 INBOX/ai/caps-escape-design.md

@@ -0,0 +1,39 @@
+# Caps Escape Design
+
+## Facts
+
+- Commit `250b0b71f` removed `services.xserver.xkb.options` from the shared host module.
+- The removed option was `caps:escape_shifted_capslock`.
+- That XKB option maps Caps Lock to Escape.
+- Shift+Caps Lock still enables Caps Lock.
+- The shared host module configures Neogaia, Zeus, and Pikachu.
+- Zeus has QMK firmware that provides equivalent keyboard behavior.
+- Neogaia and Zeus enable `modules.host.cachyos` and `modules.desktop`.
+- Pikachu enables neither module.
+- `configs/nixos/hosts/neogaia/default.nix` directly sets Neogaia-specific hardware, networking, desktop application, and Waybar settings.
+- The current configuration has no Caps Lock or Escape remapping.
+- This session runs on Zeus, not Neogaia.
+
+## Design Tree
+
+- Restore Caps Lock behavior
+  - Scope the configuration
+    - Configure Neogaia only. ✓
+  - Select the mapping behavior.
+    - Restore `caps:escape_shifted_capslock`. ✓
+    - Set `services.xserver.xkb.options` in `configs/nixos/hosts/neogaia/default.nix`. ✓
+  - Define validation for Neogaia.
+    - Run static checks and manually verify both keys after deployment. ✓
+    - Stop after source and static validation. ✓
+
+## Settled Decisions
+
+1. Configure the mapping on Neogaia only.
+2. Restore `caps:escape_shifted_capslock`.
+3. Set `services.xserver.xkb.options` directly in Neogaia's host configuration.
+4. Run static checks and manually verify both keys on Neogaia after deployment.
+5. Stop after source and static validation.
+
+## Frontier
+
+None.

docs/00 INBOX/ai/nixos-agents-style-scope-design-tree.md

@@ -0,0 +1,130 @@
+# NixOS AGENTS and style-skill scope design tree
+
+This tree reviews `configs/nixos/AGENTS.md`.
+It decides which instructions belong in Nix style rules.
+
+## Known facts
+
+- `configs/nixos/AGENTS.md` has 70 lines.
+- Each frontier question quotes the complete instruction under review.
+- This review decides instruction placement only.
+- It supplies guidance for desktop, Firefox, Neovim, AI tooling, and NixOS gotchas.
+- The `/review` prompt loads `nix-style` for Nix files.
+- `nix-style` loads every rule under `src/skills/style/nix/rules/`.
+- The current Nix rules cover Nix syntax and flake structure.
+- `configs/nixos/AGENTS.md` includes a guest-module ownership rule.
+- The RomM session missed that ownership rule during implementation.
+- `registry.window-icon` exists in `configs/nixos/registry/window-icon.nix`.
+- Desktop application modules declare window classes and rendered icons in that registry.
+- Hyprland reads the registry to generate workspace icon mappings.
+- The AGENTS instruction named the registry after its introduction on 2026-10-11.
+- The `class` option describes its window-class match.
+- The `icon` option describes a string icon value.
+- The registry schema does not document the rendered-string requirement or the class-discovery command.
+- Hyprland renders `config.registry.env` as its `env` configuration.
+- The `env` registry replaced direct Hyprland environment declarations in commit `370cab8f8`.
+- `configs/nixos/modules/desktop/apps/r2modman.nix` enables a graphical application without a `registry.window-icon` declaration.
+- The current registry modules already validate value uniqueness and field combinations.
+- They do not validate required producer declarations.
+- The registry collector is specific to the NixOS dotfiles configuration.
+- A non-discoverable environment fact records local behavior that configuration, source code, and normal tool help do not reveal.
+- The current instruction groups are:
+  - Desktop: window icons, Hyprland environment variables, and cursor settings.
+  - Firefox: search-engine declarations and extension-storage compatibility.
+  - Neovim: Nixvim package-set selection.
+  - AI tooling: Herdr key names and Pi configuration locations.
+  - Gotchas: registry APIs, guest boundaries, SOPS ownership, guest mounts, deployment, DNS, hardware, module paths, and host commands.
+
+## Decision tree
+
+- Instruction review
+  - Settled: Review each complete instruction in source order.
+  - Registry conventions
+    - Settled: Create `src/skills/style/nix/rules/dotfiles.md`.
+    - Graphical application icon instruction
+      - Settled: Move the complete instruction to `src/skills/style/nix/rules/dotfiles.md`.
+    - Hyprland session environment instruction
+      - Settled: Remove it because `registry.env` superseded it.
+    - Firefox search-engine instruction
+      - Settled: Remove it because the Firefox module documents the configuration.
+    - Firefox extension-storage instruction
+      - Settled: Move it to a comment above `extensions.settings` in `modules/desktop/apps/firefox.nix`.
+    - Neovim package-set instruction
+      - Settled: Move it to a comment above `nixpkgs.useGlobalPackages` in `modules/shell/nvim/default.nix`.
+    - Herdr key-name instruction
+      - Settled: Remove it.
+    - Pi extension location instruction
+      - Settled: Remove it. It recorded a completed migration from NixOS-managed sources.
+    - Pi generated-settings instruction
+      - Settled: Remove it because the Pi module documents the managed boundary.
+    - Library-function qualification instruction
+      - Settled: Generalize it and move it to `src/skills/style/nix/rules/general.md`.
+    - Guest ownership instruction
+      - Settled: Move it to `src/skills/style/nix/rules/dotfiles.md`.
+    - SOPS guest-owner instruction
+      - Settled: Move it to `src/skills/style/nix/rules/dotfiles.md`.
+    - Nested guest-mount instruction
+      - Settled: Remove it because the example describes no active layout.
+    - `deploy-to` instruction
+      - Settled: Remove it. Agents cannot complete its remote sudo steps.
+    - Pikachu deploy-closure instruction
+      - Settled: Remove it because `deploy-to` encodes the required import path.
+    - Public-DNS instruction
+      - Settled: Remove it.
+    - Hardware-configuration instruction
+      - Settled: Remove it because the root auto-generated-file rule owns the policy.
+    - Chaotic-cache instruction
+      - Settled: Remove it because `modules.host.cachyos.enable` owns the cache configuration.
+    - Module-relative-path instruction
+      - Settled: Remove it because Nix evaluation reports invalid paths.
+    - Pikachu `uptime` instruction
+      - Settled: Remove it because the command exposes its supported options.
+    - Guest-ACME-evaluation instruction
+      - Settled: Remove it because current direct guest evaluation succeeds.
+    - systemd executable instruction
+      - Settled: Remove it because `general.md` already requires `lib.getExe'` for non-main executables.
+
+## Frontier
+
+The frontier is empty.
+
+## Shared understanding
+
+Remove these instructions from `configs/nixos/AGENTS.md`:
+
+- Hyprland session environment
+- Firefox search engines
+- Herdr shifted punctuation
+- Pi extension source location
+- Pi managed settings
+- Nested guest mounts
+- `deploy-to` behavior
+- Pikachu deploy closures
+- Public DNS boundary
+- Hardware configuration
+- Chaotic cache selection
+- Module-relative paths
+- Pikachu boot time
+- Guest ACME evaluation
+- systemd executable selection
+
+Move these instructions intact to `src/skills/style/nix/rules/dotfiles.md`:
+
+- Graphical application icons
+- Guest ownership
+- SOPS guest ownership
+
+Move the generalized library-function qualification convention to
+`src/skills/style/nix/rules/general.md`.
+
+Move these instructions to code comments:
+
+- Firefox extension storage: above `extensions.settings` in
+  `configs/nixos/modules/desktop/apps/firefox.nix`
+- Nixvim package set: above `nixpkgs.useGlobalPackages` in
+  `configs/nixos/modules/shell/nvim/default.nix`
+
+The review does not decide how to restore the auto-generated status of
+`configs/nixos/hosts/pikachu/hardware-configuration.nix`.
+
+The operator confirmed this design and authorized the edits.

docs/00 INBOX/ai/romm-migration-design-tree.md

@@ -0,0 +1,106 @@
+# RomM migration design tree
+
+## Known facts
+
+- RomM is listed under the Pikachu and Pichu Proxmox migration.
+- Pikachu already runs VLAN 20 guests with persistent storage.
+- Nixpkgs provides a tested `services.romm` module at version 5.2.0.
+- `services.romm.dataDir` defaults to `/var/lib/romm`.
+- RomM stores the game library below `/var/lib/romm/library`.
+- RomM stores saves, states, user-uploaded screenshots, and profile avatars below `/var/lib/romm/assets/users/<user-id>/`.
+- RomM derives `<user-id>` from its database ID, so it is not a stable configured username.
+- RomM stores scraped covers and provider-sourced game screenshots below `/var/lib/romm/resources/roms/<platform-id>/<rom-id>/`.
+- RomM 5.4.0 exists upstream but Nixpkgs does not package it.
+- RomM 5.3.0 requires an explicit library-structure declaration in `config.yml`.
+- The 5.4.0 package phase must add the Structure A declaration for `roms/{platform}/{game}` and `bios/{platform}`.
+- RomM 5.4.0 can modify ROM files in place only when library conversion is explicitly enabled.
+- RomM 5.3.0 identifies hashed ROMs by CRC, MD5, and SHA-1 after a scan.
+- RomM 5.3.0 matches moved or renamed hashed ROMs to their existing database records.
+- This migration is a fresh installation.
+- The new library will receive selected ROMs from the old installation.
+- Copyparty serves `/pikachu/media` as its writable `/media` volume.
+
+## Decision tree
+
+- RomM deployment
+  - Target location
+    - Settled: Deploy a dedicated guest on Pikachu.
+    - Package source
+      - Settled: Start with Nixpkgs's RomM 5.2.0 package.
+      - Settled: Package RomM 5.4.0 locally as follow-up work.
+    - Persistent storage
+      - Settled: Use `/var/lib/romm` in the guest.
+      - Settled: Mount `/pikachu/data/romm` at `/var/lib/romm`.
+      - Settled: Mount the ROM library separately at `/var/lib/romm/library`.
+      - Settled: Mount user assets separately for Copyparty browsing.
+      - Settled: Keep scraped resources in `/pikachu/data/romm`.
+      - Settled: Mount `/pikachu/media/games/emulation/roms` at `/var/lib/romm/library/roms`.
+      - Settled: Mount `/pikachu/media/games/emulation/bios` at `/var/lib/romm/library/bios`.
+      - Settled: Mount `/pikachu/media/games/emulation/users` at `/var/lib/romm/assets/users`.
+      - Settled: Mount `/pikachu/data/romm/postgresql` at `/var/lib/postgresql`.
+      - Settled: Mount `/pikachu/data/romm/redis` at `/var/lib/redis-romm`.
+    - Guest networking
+      - Settled: Follow the standard guest-provisioning protocol, including its VLAN 20 DHCP reservation.
+    - Public route
+      - Settled: Retain the old installation's public domain.
+    - Metadata and authentication secrets
+      - Settled: Configure IGDB, ScreenScraper, and SteamGridDB.
+      - Settled: Use RomM's default metadata and artwork priority.
+      - Settled: Manage the 5.4.0 `config.yml` declaratively with Nix.
+      - Settled: Use RomM's default region and English-language priority.
+    - Save synchronization
+      - Settled: Disable automatic browser save synchronization in 5.4.0.
+    - Authentication
+      - Settled: Use RomM's built-in accounts initially.
+      - Deferred: Configure RomM as an OpenID Connect client after an identity provider exists.
+    - ROM selection and import
+      - Settled: The operator manually selects and copies ROMs after deployment.
+      - Settled: Do not create automatic import tooling or a ROM manifest.
+    - Capacity and service limits
+      - Settled: Limit the guest to `4G` memory, `200%` CPU, and `512` tasks.
+    - Library scanning
+      - Settled: Enable the filesystem watcher for library changes.
+    - Cutover and retirement
+      - Out of scope: The operator manages the old installation and public-domain cutover outside this monorepo.
+  - Work phases
+    - Settled: Create a RomM guest module with Nixpkgs's package in [[0014-romm-guest]].
+    - Settled: Package RomM 5.4.0 locally and reuse `services.romm` in [[0015-romm-5-4-package]].
+    - Settled: Move selected ROMs after the service deployment is ready in [[0016-romm-manual-rom-migration]].
+
+## Settled decisions
+
+- **Migration model:** Start with a fresh RomM installation.
+- **Data scope:** Copy selected ROMs from the old installation.
+- **Old application state:** Do not migrate the database, users, sessions, or job queue.
+- **Deployment target:** Deploy a dedicated guest on Pikachu.
+- **Initial package:** Use Nixpkgs's RomM 5.2.0 package.
+- **Package follow-up:** Package RomM 5.4.0 locally.
+- **Work phases:** Deploy the guest, update the package, then import selected ROMs.
+- **Guest data directory:** Use RomM's default `/var/lib/romm`.
+- **Host application data:** Mount `/pikachu/data/romm` at `/var/lib/romm`.
+- **ROM library:** Mount it separately at `/var/lib/romm/library`.
+- **Storage classification:** Preserve the distinction between application data and browsable media.
+- **User assets:** Mount `/pikachu/media/games/emulation/users` at `/var/lib/romm/assets/users` to expose saves, states, user-uploaded screenshots, and profile avatars through Copyparty.
+- **ROM paths:** Mount `/pikachu/media/games/emulation/roms` and `/pikachu/media/games/emulation/bios` at the corresponding children of `/var/lib/romm/library`.
+- **Guest networking:** Follow the standard guest-provisioning protocol.
+- **Public route:** Retain the old installation's public domain.
+- **Metadata providers:** Configure IGDB, ScreenScraper, and SteamGridDB in both package phases.
+- **Metadata priority:** Use RomM's default metadata and artwork priority.
+- **RomM configuration:** Manage the required 5.4.0 `config.yml` declaratively with Nix.
+- **Scan region and language priority:** Use RomM's defaults.
+- **Initial authentication:** Use RomM's built-in accounts.
+- **OIDC follow-up:** Track it in [[0013-oidc-provider]].
+- **ROM migration:** The operator manually selects and copies ROMs after deployment.
+- **Guest limits:** Cap the guest at `4G` memory, `200%` CPU, and `512` tasks.
+- **Library watcher:** Enable the filesystem watcher.
+- **PostgreSQL:** Mount `/pikachu/data/romm/postgresql` at `/var/lib/postgresql`.
+- **Redis:** Mount `/pikachu/data/romm/redis` at `/var/lib/redis-romm`.
+- **Browser save synchronization:** Disable it in 5.4.0 until browser emulation is needed.
+
+## Frontier
+
+No in-scope decisions remain.
+
+## Next question
+
+None. Present the settled design for confirmation before implementation.

docs/00 INBOX/ai/tldr-design-tree.md

@@ -0,0 +1,36 @@
+# TL;DR command design tree
+
+This tree reviews the implementation decisions in `tldr-design.md`.
+
+## Settled decisions
+
+- **Command form and scope:** Add the project prompt template `.pi/prompts/tldr.md`.
+- **Explicit source interface:** Interpret command arguments as a free-form source request.
+- **Default source:** Summarize the preceding agent response when no source request exists.
+- **Summary shape:** Produce a compact summary without a fixed format.
+- **Unavailable or unclear sources:** Add no special handling policy.
+
+## Decision tree
+
+- Command behavior
+  - Source selection
+    - Settled: Use the preceding agent response without arguments.
+    - Settled: Interpret arguments as a free-form source request.
+  - Output
+    - Settled: Produce a compact summary without a fixed format.
+    - Summary language
+      - Settled: Do not prescribe a language policy.
+  - Error behavior
+    - Settled: Add no special handling policy for unavailable or unclear sources.
+
+## Frontier
+
+The frontier is empty.
+
+## Shared understanding pending confirmation
+
+The project gets a `/tldr` prompt command in `.pi/prompts/tldr.md`.
+Without arguments, it summarizes the preceding agent response.
+With arguments, it interprets them as a free-form source request.
+It produces a compact summary with no fixed format or language policy.
+The prompt adds no special policy for unavailable or unclear sources.

docs/00 INBOX/ai/tldr-design.md

@@ -0,0 +1,39 @@
+# TL;DR command design
+
+## Goal
+
+Add a Pi command that summarizes the preceding agent response by default.
+An explicit command argument identifies the source to summarize.
+
+## Decision tree
+
+- Command form and scope
+  - Settled: Create a project prompt template at `.pi/prompts/tldr.md`.
+  - Command interface
+    - Settled: Pass arguments through as a free-form source request.
+    - Default source selection
+      - Settled: Summarize the preceding agent response when no source request exists.
+      - Summary format and length
+        - Settled: Use a compact, format-free summary.
+        - Ambiguous-source behavior
+          - Settled: Add no special handling policy.
+
+## Settled decisions
+
+- **Q1 — Command form and scope:** Create a project prompt template.
+- **Q2 — Explicit source interface:** Accept a free-form source request.
+- **Default source selection:** Summarize the preceding agent response when the command has no arguments.
+- **Q3 — Summary shape:** Use a compact, format-free summary.
+- **Q4 — Unavailable source:** Add no special handling policy.
+
+## Frontier
+
+The frontier is empty.
+
+## Shared understanding pending confirmation
+
+The project gets a `/tldr` prompt command in `.pi/prompts/tldr.md`.
+Without arguments, it summarizes the preceding agent response.
+With arguments, it interprets them as a free-form source request.
+It produces a compact summary with no fixed format.
+The prompt adds no special policy for unavailable or unclear sources.

docs/00 INBOX/credits.md

@@ -0,0 +1,14 @@
+---
+aliases:
+  - Credits
+date-created: 2026-10-17 18:08
+date-modified: 2026-10-17 23:33
+---
+
+# Credits
+
+| Path                                | Original                                                                                | Notes |   |
+|-------------------------------------|-----------------------------------------------------------------------------------------|-------|---|
+| `//src/skills/agent-writing-skills` | <https://github.com/mattpocock/skills/tree/main/skills/productivity/writing-for-agents> |       |   |
+| `//src/skills/retrospect`           | <https://github.com/mattpocock/skills/tree/main/skills/engineering/retro>               |       |   |
+|                                     |                                                                                         |       |   |

docs/03 wrk/monorepo.md

@@ -0,0 +1,8 @@
+---
+aliases:
+  - Monorepo
+date-created: 2026-10-14 22:51
+date-modified: 2026-10-14 22:54
+---
+
+# Monorepo

docs/03 wrk/tasks/0001-todo-cli.md

@@ -3,10 +3,11 @@ aliases:
   - "0001"
   - TODO CLI
 date-created: 2026-10-14 00:49
-date-modified: 2026-10-14 16:33
+date-modified: 2026-10-14 22:52
 status: open
 area: "[[productivity]]"
-blocked-by: []
+blocked-by:
+  - "[[0005-rust-cli-scaffolding]]"
 subtasks: []
 ---
 

docs/03 wrk/tasks/0002-nixos-refactor.md

@@ -3,8 +3,8 @@ aliases:
   - "0002"
   - NixOS Refactor
 date-created: 2026-10-14 16:07
-date-modified: 2026-10-14 18:17
+date-modified: 2026-10-19 16:44
-status: open
+status: closed
 area: "[[dotfiles]]"
 blocked-by: []
 subtasks: []
@@ -64,40 +64,40 @@ Complete the remaining Nix refactor work.
 - [x] `configs/nixos/modules/desktop/utilities/userdirs.nix`
 - [x] `configs/nixos/modules/services/actualbudget.nix`
 - [x] `configs/nixos/modules/services/copyparty.nix`
-- [ ] `configs/nixos/modules/services/homepage.nix`
+- [x] `configs/nixos/modules/services/homepage.nix`
-- [ ] `configs/nixos/modules/services/navidrome.nix`
+- [x] `configs/nixos/modules/services/navidrome.nix`
-- [ ] `configs/nixos/modules/services/reverse-proxy.nix`
+- [x] `configs/nixos/modules/services/reverse-proxy.nix`
-- [ ] `configs/nixos/modules/services/steamcmd.nix`
+- [x] `configs/nixos/modules/services/steamcmd.nix`
-- [ ] `configs/nixos/modules/services/games/palworld.nix`
+- [x] `configs/nixos/modules/services/games/palworld.nix`
-- [ ] `configs/nixos/modules/services/games/valheim.nix`
+- [x] `configs/nixos/modules/services/games/valheim.nix`
-- [ ] `configs/nixos/modules/shell/bazel.nix`
+- [x] `configs/nixos/modules/shell/bazel.nix`
-- [ ] `configs/nixos/modules/shell/direnv/default.nix`
+- [x] `configs/nixos/modules/shell/direnv/default.nix`
-- [ ] `configs/nixos/modules/shell/fish/default.nix`
+- [x] `configs/nixos/modules/shell/fish/default.nix`
-- [ ] `configs/nixos/modules/shell/gitea-axi.nix`
+- [x] `configs/nixos/modules/shell/gitea-axi.nix`
-- [ ] `configs/nixos/modules/shell/git.nix`
+- [x] `configs/nixos/modules/shell/git.nix`
-- [ ] `configs/nixos/modules/shell/herdr.nix`
+- [x] `configs/nixos/modules/shell/herdr.nix`
-- [ ] `configs/nixos/modules/shell/nvim/default.nix`
+- [x] `configs/nixos/modules/shell/nvim/default.nix`
-- [ ] `configs/nixos/modules/shell/omp.nix`
+- [x] `configs/nixos/modules/shell/omp.nix`
-- [ ] `configs/nixos/modules/shell/pi/default.nix`
+- [x] `configs/nixos/modules/shell/pi/default.nix`
-- [ ] `configs/nixos/modules/shell/python.nix`
+- [x] `configs/nixos/modules/shell/python.nix`
-- [ ] `configs/nixos/modules/shell/rust.nix`
+- [x] `configs/nixos/modules/shell/rust.nix`
-- [ ] `configs/nixos/modules/shell/tmux/default.nix`
+- [x] `configs/nixos/modules/shell/tmux/default.nix`
-- [ ] `configs/nixos/modules/shell/yazi.nix`
+- [x] `configs/nixos/modules/shell/yazi.nix`
-- [ ] `src/dotfiles/guests.nix`
+- [x] `src/dotfiles/guests.nix`
 - [x] `configs/nixos/guests/actualbudget.nix`
-- [ ] `configs/nixos/guests/copyparty.nix`
+- [x] `configs/nixos/guests/copyparty.nix`
-- [ ] `configs/nixos/guests/homepage.nix`
+- [x] `configs/nixos/guests/homepage.nix`
-- [ ] `configs/nixos/guests/navidrome.nix`
+- [x] `configs/nixos/guests/navidrome.nix`
-- [ ] `configs/nixos/guests/palworld.nix`
+- [x] `configs/nixos/guests/palworld.nix`
-- [ ] `configs/nixos/guests/reverse-proxy.nix`
+- [x] `configs/nixos/guests/reverse-proxy.nix`
-- [ ] `configs/nixos/guests/valheim.nix`
+- [x] `configs/nixos/guests/valheim.nix`
-- [ ] `src/dotfiles/hosts.nix`
+- [x] `src/dotfiles/hosts.nix`
-- [ ] `configs/nixos/hosts/neogaia/default.nix`
+- [x] `configs/nixos/hosts/neogaia/default.nix`
-- [ ] `configs/nixos/hosts/neogaia/disk.nix`
+- [x] `configs/nixos/hosts/neogaia/disk.nix`
-- [ ] `configs/nixos/hosts/neogaia/hardware-configuration.nix`
+- [x] `configs/nixos/hosts/neogaia/hardware-configuration.nix`
-- [ ] `configs/nixos/hosts/pikachu/default.nix`
+- [x] `configs/nixos/hosts/pikachu/default.nix`
-- [ ] `configs/nixos/hosts/pikachu/disk.nix`
+- [x] `configs/nixos/hosts/pikachu/disk.nix`
-- [ ] `configs/nixos/hosts/pikachu/hardware-configuration.nix`
+- [x] `configs/nixos/hosts/pikachu/hardware-configuration.nix`
-- [ ] `configs/nixos/hosts/zeus/default.nix`
+- [x] `configs/nixos/hosts/zeus/default.nix`
-- [ ] `configs/nixos/hosts/zeus/disk.nix`
+- [x] `configs/nixos/hosts/zeus/disk.nix`
-- [ ] `configs/nixos/hosts/zeus/hardware-configuration.nix`
+- [x] `configs/nixos/hosts/zeus/hardware-configuration.nix`

docs/03 wrk/tasks/0004-dictionary-cli.md

@@ -0,0 +1,18 @@
+---
+aliases:
+  - "0004"
+  - Dictionary CLI
+date-created: 2026-10-14 22:47
+date-modified: 2026-10-14 22:57
+status: open
+area: "[[productivity]]"
+blocked-by:
+  - "[[0005-rust-cli-scaffolding]]"
+subtasks: []
+---
+
+# Dictionary CLI
+
+A CLI tool that basically works as a strict allowlist for words used in prose such as markdown files or strings passed to the tool.
+
+The basic idea can be found in [[dictionary]], but as it would be a deterministic tool, a strict allowlist might be more beneficial.

docs/03 wrk/tasks/0005-rust-cli-scaffolding.md

@@ -0,0 +1,21 @@
+---
+aliases:
+  - "0005"
+  - Rust CLI Scaffolding
+date-created: 2026-10-14 22:50
+date-modified: 2026-10-15 23:48
+# Allowed: open, closed
+status: open
+area: "[[monorepo]]"
+blocked-by:
+  - "[[0009-bazel-project-scaffolding]]"
+subtasks: []
+---
+
+# Rust CLI Scaffolding
+
+One of the biggest patterns I noticed when working on Rust projects in the monorepo was that they were mostly CLI tools. I found myself doing the same boilerplate over and over again.
+
+I think much like Google, having a function that handles sensible defaults would be a good idea.
+
+But also, maybe a Rust project set up to avoid the common pitfalls?

docs/03 wrk/tasks/0006-wallpaper-randomizer.md

@@ -0,0 +1,23 @@
+---
+aliases:
+  - "0006"
+  - Wallpaper Randomizer
+date-created: 2026-10-14 23:00
+date-modified: 2026-10-14 23:05
+status: open
+area: "[[dotfiles]]"
+blocked-by:
+  - "[[0005-rust-cli-scaffolding]]"
+subtasks: []
+---
+
+# Wallpaper Randomizer
+
+This is kind of two ideas in one:
+
+1. Randomizing the desktop wallpaper from a select list.
+2. A web service endpoint that provides a random image from a configured directory/set.
+
+The first idea might not need any special code, so it wouldn't be a CLI.
+
+I think that any tools I write should have both a non-interactive mode and a TUI or web server mode.

docs/03 wrk/tasks/0007-obsidian-source-notes.md

@@ -0,0 +1,20 @@
+---
+aliases:
+  - "0007"
+  - Obsidian Source Notes
+date-created: 2026-10-14 23:11
+date-modified: 2026-10-14 23:14
+status: open
+area: "[[productivity]]"
+blocked-by: []
+subtasks: []
+---
+
+# Obsidian Source Notes
+
+This is more of a note to collect ideas for `ingest-*` skills.
+
+## Skill Ideas
+
+- `ingest-youtube-transcript`
+- `ingest-youtube-video`

docs/03 wrk/tasks/0008-vim-obsidian-markdown-renderer.md

@@ -0,0 +1,15 @@
+---
+aliases:
+  - "0008"
+  - Vim Obsidian Markdown Renderer
+date-created: 2026-10-14 23:18
+date-modified: 2026-10-14 23:20
+status: open
+area: "[[dotfiles]]"
+blocked-by: []
+subtasks: []
+---
+
+# Vim Obsidian Markdown Renderer
+
+The current plugin I use for Markdown rendering in Vim has some issues, but most importantly it doesn't have full feature parity with Obsidian. I think from the top of my head, wikilinks can't be followed for example.

docs/03 wrk/tasks/0009-bazel-project-scaffolding.md

@@ -0,0 +1,17 @@
+---
+aliases:
+  - "0009"
+  - Bazel Project Scaffolding
+date-created: 2026-10-15 23:46
+date-modified: 2026-10-16 00:32
+status: open
+area: "[[monorepo]]"
+blocked-by: []
+subtasks: []
+---
+
+# Bazel Project Scaffolding
+
+A small utility library that sets up projects when needed.
+
+The first goal would be to scaffold a basic Rust project with a library, tests and binary to avoid manual toil.

docs/03 wrk/tasks/0010-fix-main-monitor-detection.md

@@ -0,0 +1,18 @@
+---
+aliases:
+  - "0010"
+  - Fix Main Monitor Detection
+date-created: 2026-10-17 09:24
+date-modified: 2026-10-17 09:25
+# Allowed: open, closed
+status: open
+area: "[[dotfiles]]"
+blocked-by: []
+subtasks: []
+---
+
+# Fix Main Monitor Detection
+
+Most of the time, games detect the HDMI monitor as the main monitor and show the wrong resolution options and/or have display issues.
+
+The HDMI monitor is a vertical monitor with 1080x1920 resolution so that leads to a lot of problems I can't fix from within the game because it gets cut off.

docs/03 wrk/tasks/0011-gitea-axi-rust-migration.md

@@ -0,0 +1,20 @@
+---
+aliases:
+  - "0011"
+  - Gitea Axi Rust Migration
+date-created: 2026-10-18 15:43
+date-modified: 2026-10-19 16:46
+status: open
+area: "[[monorepo]]"
+blocked-by:
+  - "[[0005-rust-cli-scaffolding]]"
+subtasks: []
+---
+
+# Gitea Axi Rust Migration
+
+Rewrite `gitea-axi` in Rust and migrate it into this monorepo.
+
+Preserve its Gitea workflow and command behavior.
+
+Replace the external Home Manager module in `configs/nixos/modules/shell/gitea-axi.nix`.

docs/03 wrk/tasks/0012-gitea-axi-pr-merge-guardrail.md

@@ -0,0 +1,19 @@
+---
+aliases:
+  - "0012"
+  - Gitea Axi PR Merge Guardrail
+date-created: 2026-10-18 15:43
+date-modified: 2026-10-18 15:43
+status: open
+area: "[[monorepo]]"
+blocked-by:
+  - "[[0011-gitea-axi-rust-migration]]"
+subtasks: []
+---
+
+# Gitea Axi PR Merge Guardrail
+
+Guard `gitea-axi pr merge` against an unpushed local rewrite.
+
+Compare the checked-out matching branch with the remote pull-request head.
+Require explicit confirmation when local commits are absent from that head.

docs/03 wrk/tasks/0013-oidc-provider.md

@@ -0,0 +1,17 @@
+---
+aliases:
+  - "0013"
+  - OIDC Provider
+date-created: 2026-10-19 19:40
+date-modified: 2026-10-19 19:40
+status: open
+area: "[[homelab]]"
+blocked-by: []
+subtasks: []
+---
+
+# OIDC Provider
+
+Deploy an OpenID Connect provider for homelab services.
+
+Configure a RomM client after the provider is available.

docs/03 wrk/tasks/0014-romm-guest.md

@@ -0,0 +1,42 @@
+---
+aliases:
+  - "0014"
+  - RomM Guest
+date-created: 2026-10-19 19:40
+date-modified: 2026-10-19 19:40
+status: closed
+area: "[[homelab]]"
+blocked-by: []
+subtasks: []
+---
+
+# RomM Guest
+
+Deploy a dedicated RomM guest on Pikachu with Nixpkgs RomM 5.2.0.
+
+Follow the standard guest-provisioning protocol.
+Keep the old installation's public domain.
+
+Use the local PostgreSQL and Redis instances from `services.romm`.
+Set guest limits to `4G` memory, `200%` CPU, and `512` tasks.
+Enable the library watcher.
+Use RomM built-in accounts.
+
+Mount persistent application data:
+
+```text
+/pikachu/data/romm             -> /var/lib/romm
+/pikachu/data/romm/postgresql  -> /var/lib/postgresql
+/pikachu/data/romm/redis       -> /var/lib/redis-romm
+```
+
+Mount browsable media:
+
+```text
+/pikachu/media/games/emulation/roms  -> /var/lib/romm/library/roms
+/pikachu/media/games/emulation/bios  -> /var/lib/romm/library/bios
+/pikachu/media/games/emulation/users -> /var/lib/romm/assets/users
+```
+
+Configure IGDB, ScreenScraper, and SteamGridDB from secrets.
+Use RomM's default metadata, artwork, region, and language priorities.

docs/03 wrk/tasks/0015-romm-5-4-package.md

@@ -0,0 +1,31 @@
+---
+aliases:
+  - "0015"
+  - RomM 5.4 Package
+date-created: 2026-10-19 19:40
+date-modified: 2026-10-19 19:40
+status: open
+area: "[[homelab]]"
+blocked-by:
+  - "[[0014-romm-guest]]"
+subtasks: []
+---
+
+# RomM 5.4 Package
+
+Package RomM 5.4.0 locally.
+Reuse `services.romm`.
+
+Manage `config.yml` declaratively with Nix.
+Declare the Structure A library layout:
+
+```yaml
+filesystem:
+  structure:
+    default: "roms/{platform}/{game}"
+    firmware: "bios/{platform}"
+```
+
+Keep library conversion disabled.
+Disable automatic browser save synchronization until browser emulation is needed.
+Keep the IGDB, ScreenScraper, and SteamGridDB configuration unchanged.

docs/03 wrk/tasks/0016-romm-manual-rom-migration.md

@@ -0,0 +1,24 @@
+---
+aliases:
+  - "0016"
+  - RomM Manual ROM Migration
+date-created: 2026-10-19 19:40
+date-modified: 2026-10-19 19:40
+status: open
+area: "[[homelab]]"
+blocked-by:
+  - "[[0014-romm-guest]]"
+subtasks: []
+---
+
+# RomM Manual ROM Migration
+
+Manually copy selected ROMs and BIOS files from the old installation.
+
+Copy ROMs to `/pikachu/media/games/emulation/roms`.
+Copy BIOS files to `/pikachu/media/games/emulation/bios`.
+
+Do not migrate the old database, users, sessions, job queue, or other application state.
+Do not add automatic import tooling or a ROM manifest.
+
+This task does not change monorepo configuration.

docs/99 meta/templates/task.md

@@ -1,6 +1,7 @@
 ---
 date-created: "{{date}} {{time}}"
-status: open # Allowed: open, closed
+# Allowed: open, closed
+status: open
 area:
 blocked-by:
 -

docs/TODO.md

@@ -12,7 +12,6 @@
 # Project Ideas
 
 - onepacerr (requires raichu migration)
-- obsidian-source-notes (`ingest-*` skills)
 - obsidian-game-backlog
 - breadcrumbs
 - gingko
@@ -20,10 +19,7 @@
 - home assistant agent integration
 - wallpaper randomizer
 - ebook tool (might be an overlap with `ingest-ebook`)
-- vim-obsidian-markdown
-- dictionary cli tool
 - kitchen.md
-- Rust project creator
 - random number generator utility
 - reading material helper
 - valheim progression DSL + graph
@@ -32,9 +28,7 @@
 
 # Migration
 
-- [[dictionary]] -> cli tool
 - [[technical-writing]] -> skill?
-- gitea-axi
 - skills
   - claude-code/codebase-design
   - claude-code/implement

flake.lock

@@ -75,11 +75,11 @@
         "nixpkgs": "nixpkgs"
       },
       "locked": {
-        "lastModified": 1789232619,
+        "lastModified": 1792048969,
-        "narHash": "sha256-8pyRIm67V5dtVN3A4Mu0e2a7DIhUhWp6y3NSVS6WBwU=",
+        "narHash": "sha256-+Scy7etDouWzvEV9d6zEDz/TxOoGzfSJau9+yp/zWM4=",
         "owner": "chaotic-cx",
         "repo": "nyx",
-        "rev": "0265d6932168aef12c47f03a565786410ba877e0",
+        "rev": "c283697bfcc3770232139b86c4ad19fd8a62cd6f",
         "type": "github"
       },
       "original": {
@@ -96,11 +96,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1781152676,
+        "lastModified": 1789770686,
-        "narHash": "sha256-RxWs5ND31KzTG7wvMM+PMfUjyNpmIEr999lqNARaM5o=",
+        "narHash": "sha256-uZkBR7yHdIKUFB5SZdfgh1qkGfI3XmYmI/lTiquxbck=",
         "owner": "nix-community",
         "repo": "disko",
-        "rev": "ff8702b4de27f72b4c78573dfb89ec74e36abdf1",
+        "rev": "725ea35e410ad83be4931d1bff7e090eacaf3563",
         "type": "github"
       },
       "original": {
@@ -117,11 +117,11 @@
       },
       "locked": {
         "dir": "pkgs/firefox-addons",
-        "lastModified": 1789185767,
+        "lastModified": 1792036980,
-        "narHash": "sha256-jaWL9beN3ifjNzNEBxo0abA+TSi0n7Q8LRQvZTxEqaM=",
+        "narHash": "sha256-y28wLbzsIbBDeToXHhAdoTs+lcH/d7/1Pky40N6/CoI=",
         "owner": "rycee",
         "repo": "nur-expressions",
-        "rev": "811170bb74b11f3b9489618ce77487cc58c79067",
+        "rev": "cbf0380669792dd8f6dbb891d3be33aae8fa4625",
         "type": "gitlab"
       },
       "original": {
@@ -155,11 +155,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1788450739,
+        "lastModified": 1790898470,
-        "narHash": "sha256-glZLQlzIn1fXH6PazR2iUmTo7kzzyYSshrWhLS9TqCU=",
+        "narHash": "sha256-zTwuwezD0w3hpga6a6P8emidzAZFyWqlNeRBlpWOOl8=",
         "owner": "hercules-ci",
         "repo": "flake-parts",
-        "rev": "31729ca8cbdb4fa927b34e5f4353e6a83f39e993",
+        "rev": "024633cd702b10285db5cb19b40ad48d2399ba60",
         "type": "github"
       },
       "original": {
@@ -267,11 +267,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1789183968,
+        "lastModified": 1791940130,
-        "narHash": "sha256-pEWnYdIF1pBMZwarp/rEPzl+Lknhm5hhjSfeLxH8nAA=",
+        "narHash": "sha256-hRCCi4fp5IjLz4cA8+RkUsaYAHzbd7M5FSAjKOq4E+k=",
         "owner": "nix-community",
         "repo": "home-manager",
-        "rev": "cd1c9e552f41894aeb5cc5cb353d5a1d61550357",
+        "rev": "92a0ed1be12f1e20093faf8af63bae2c74ad521e",
         "type": "github"
       },
       "original": {
@@ -308,11 +308,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1789267719,
+        "lastModified": 1792098797,
-        "narHash": "sha256-FIlwGdFjtpGYyL8g/RZZvIxshT9gdwJfcMjHRWHrc4U=",
+        "narHash": "sha256-u4DIIZgznEGzByxXrxWCXK19VcMtTBmi/M9gZ7Jk9bM=",
         "owner": "nix-community",
         "repo": "home-manager",
-        "rev": "10d0acd7f81e3c035b452daa39ac7d1673880729",
+        "rev": "640725978b99076e18e069f6e7d2959124584f8f",
         "type": "github"
       },
       "original": {
@@ -328,11 +328,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1789127271,
+        "lastModified": 1791721963,
-        "narHash": "sha256-vL/Fpo7Zxa7wvO8V5FAToUBvP+K3XsMB1wUglTtd/J0=",
+        "narHash": "sha256-BYhx89XJltnRyQeuDKHVL5eSqbbfAcOKuDts7QRsrxQ=",
         "owner": "NixOS",
         "repo": "nixos-hardware",
-        "rev": "24cfdc1f9344b90a1eee329a3906e2f39f4d0f1e",
+        "rev": "60a624466c08787ee8303393b7d4752d1826f14f",
         "type": "github"
       },
       "original": {
@@ -343,27 +343,24 @@
     },
     "nixpkgs": {
       "locked": {
-        "lastModified": 1789006805,
+        "lastModified": 1791734076,
-        "narHash": "sha256-xB8mKMOx1IA9vTDNLmJZ6n4wCMq/cuWBBOzGCRnqxrU=",
+        "narHash": "sha256-0dNcyVeoD61zhucIpt5BAQ+ytd6JzF641Z+A9bgltLo=",
-        "owner": "NixOS",
+        "rev": "7fd4004d0511991fff3a0524856a9be09e28c5f0",
-        "repo": "nixpkgs",
+        "type": "tarball",
-        "rev": "8ce4ef6cb6f871616146b9fe26d2a5ae594e94fe",
+        "url": "https://releases.nixos.org/nixos/unstable/nixos-26.11pre1090327.7fd4004d0511/nixexprs.tar.zst"
-        "type": "github"
       },
       "original": {
-        "owner": "NixOS",
+        "type": "tarball",
-        "ref": "nixos-unstable",
+        "url": "https://channels.nixos.org/nixos-unstable/nixexprs.tar.zst"
-        "repo": "nixpkgs",
-        "type": "github"
       }
     },
     "nixpkgs-stable": {
       "locked": {
-        "lastModified": 1789114715,
+        "lastModified": 1791883866,
-        "narHash": "sha256-ugpsyk3NM2s87vXfUiIIiibbJ4Pp0JPS5p/3mfs+q+c=",
+        "narHash": "sha256-yvkDdToRZ4IgPc9tSBzzCYJvTLf8n5eE3rgUkpCuD5A=",
         "owner": "nixos",
         "repo": "nixpkgs",
-        "rev": "21a67dc470149f337cecafbe965d8d252a390518",
+        "rev": "a033e66ab94a6bedf2c3d58d93a36319d3cbbbaf",
         "type": "github"
       },
       "original": {
@@ -375,11 +372,11 @@
     },
     "nixpkgs-unstable": {
       "locked": {
-        "lastModified": 1789073787,
+        "lastModified": 1791966044,
-        "narHash": "sha256-xfX/toC2QV707s06GbP4II/TxYF0fNQj7s5/LClNDKc=",
+        "narHash": "sha256-t1hT8UW97rFCWCJO7/pClvu0AhLN5V3uir2M0ic4w/Q=",
         "owner": "nixos",
         "repo": "nixpkgs",
-        "rev": "aff8a0b28396750446e5537a96461bc4facdb287",
+        "rev": "319e4cebef10cbb636c6ea392137b52ad105336a",
         "type": "github"
       },
       "original": {
@@ -391,11 +388,11 @@
     },
     "nixpkgs_2": {
       "locked": {
-        "lastModified": 1789149629,
+        "lastModified": 1791956154,
-        "narHash": "sha256-H6GwaZzZf+4npqv0tph94w9tZddSjFjmQrVsW0z78uk=",
+        "narHash": "sha256-jBYGPi4udQ9uo4KQp9m2J0546nJVQitG15gvZSaVVAs=",
         "owner": "nixos",
         "repo": "nixpkgs",
-        "rev": "eaad089433ca2bb662274377d33df3d0e51ef28b",
+        "rev": "55bfb4d7437bd4e09ab7cba58a43bd68eb310044",
         "type": "github"
       },
       "original": {
@@ -414,11 +411,11 @@
         "systems": "systems"
       },
       "locked": {
-        "lastModified": 1789067777,
+        "lastModified": 1792012326,
-        "narHash": "sha256-gN3OnYgJwCK0HX01dQGWusnirJuVHE9WigmSqOnyuCE=",
+        "narHash": "sha256-DApZIjaKdU4XzglZ56yrPYLhG/ajoyuQ4g4s8tkHqxU=",
         "owner": "nix-community",
         "repo": "nixvim",
-        "rev": "16baff93297c63b6d4c497f00ffcf0b844b06fc6",
+        "rev": "8485794923eb406ce3728af320392226608f724b",
         "type": "github"
       },
       "original": {
@@ -476,11 +473,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1788914643,
+        "lastModified": 1792056300,
-        "narHash": "sha256-4GuMPW90JSxXWDPUB9M+1m7fYbe3H0apOd86/zBQ2Kw=",
+        "narHash": "sha256-YU1LjDe1o2Ymg5zEiCHbmMXLkDCtxHQRiVqZ3nC5GGk=",
         "owner": "Mic92",
         "repo": "sops-nix",
-        "rev": "13616fff713a9f94055c66f15687ebdc17a335df",
+        "rev": "44931708c62937c2b1bf7953902989fa5bb9ee1b",
         "type": "github"
       },
       "original": {
@@ -509,11 +506,11 @@
         "tinted-zed": "tinted-zed"
       },
       "locked": {
-        "lastModified": 1787771653,
+        "lastModified": 1791574276,
-        "narHash": "sha256-DkkJSBOXWV/mja5Vy8az5g1KpZlsbUwlmH0BsQhowaQ=",
+        "narHash": "sha256-UMnLnZ3JlA7UTBNNNx1rNsXYoPP/cirXLCKGiQ/yDx8=",
         "owner": "danth",
         "repo": "stylix",
-        "rev": "5e3809851f486e7fc7e84b40f174c74b60ecc784",
+        "rev": "2856c4c6a845b9bec98828ef197abe176e4806bb",
         "type": "github"
       },
       "original": {
@@ -525,11 +522,11 @@
     "systems": {
       "flake": false,
       "locked": {
-        "lastModified": 1774449309,
+        "lastModified": 1791244623,
-        "narHash": "sha256-brhZ8DmuGtzkCYHJg4HEd602amKm89Y9ytsFZ5uWD1w=",
+        "narHash": "sha256-uuanbkFk8a313Zfrxdcz8Us/yQtaxNnKJtX8v0ZPpd8=",
         "owner": "nix-systems",
         "repo": "default",
-        "rev": "c29398b59d2048c4ab79345812849c9bd15e9150",
+        "rev": "d36eeed142f88b5a636f2e3856a53c61654cf29b",
         "type": "github"
       },
       "original": {
@@ -626,11 +623,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1786901030,
+        "lastModified": 1791885023,
-        "narHash": "sha256-WSFCsDSE5ffgD2MqzkM2CYjeFiKhRF/dJUN8uedb6YE=",
+        "narHash": "sha256-C3alQ2MEt0zVrUSho5NUDo3jmhbff72++bfwoqT9Omo=",
         "owner": "numtide",
         "repo": "treefmt-nix",
-        "rev": "27b3b12a8e6375f28ebe122f07d230ca5459bbfa",
+        "rev": "0d6d739a833ba9a9c93034862e4a7b061a062a1c",
         "type": "github"
       },
       "original": {

src/skills/.pi/skills/craft-skill

@@ -1 +0,0 @@
-../../craft-skill

src/skills/agent-writing-guide/SKILL-MECHANICS.md

@@ -0,0 +1,22 @@
+# Skill mechanics
+
+The skill-specific branch of [`agent-writing-guide`](SKILL.md): what changes when the document is a skill (frontmatter, the invocation choice, and router skills). Everything else about writing it is the universal reference in `SKILL.md`.
+
+## Invocation
+
+Two choices, trading the two loads:
+
+- A **model-invoked** skill keeps a `description`, so the agent can fire it autonomously, and other skills can reach it. You can still type its name: model-invocation always _includes_ user reach; a description only ever adds agent discovery, never removes the human's. The description is the skill's top-level context pointer, forced to stay loaded at all times: permanent context load in exchange for discoverability. A model-invoked skill whose content is all reference is also one home for shared reference: another skill can invoke it, so reference needed by several skills lives in one place. Mechanics: omit `disable-model-invocation`, and write a model-facing description carrying the trigger branches (the pointer-writing rules in `SKILL.md` apply in full).
+- A **user-invoked** skill strips the description from the agent's reach: only the human typing its name can invoke it, and no other skill can. Zero context load, but it spends cognitive load: you are the index that must remember it exists. Mechanics: set `disable-model-invocation: true`; the `description` becomes human-facing: a one-line summary, trigger lists stripped.
+
+Pick model-invocation only when the agent must reach the skill on its own, or another skill must. If it only ever fires by hand, make it user-invoked and pay no context load.
+
+Shared reference that two user-invoked skills both need can live in neither: neither skill exposes model invocation. Push it to a plain file outside the skill system: external reference any skill can point at.
+
+## Splitting by invocation
+
+The invocation cut of splitting (the sequence cut lives in `SKILL.md`): split off a model-invoked skill when you have a distinct leading word that should trigger it on its own (a trigger word you actually use in your prompts), or another skill must reach it. You pay context load for the new always-loaded description, so that independent reach has to be worth it.
+
+## Router skills
+
+When user-invoked skills multiply past what you can remember, that piled-up cognitive load is cured by a **router skill**: one user-invoked skill that names the others and when to reach for each, so the human has one skill to remember instead of many. It can only hint, never fire them: `disable-model-invocation: true` prevents model invocation, so nothing but the human can reach them.

src/skills/agent-writing-guide/SKILL.md

@@ -0,0 +1,81 @@
+---
+name: agent-writing-guide
+description: Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
+---
+
+Reference for writing any document an agent consumes: a skill, an `AGENTS.md` / `CLAUDE.md`, a doc reached by a pointer. The packaging differs; the writing does not: the same levers make each one predictable, since the agent takes the same _process_ every run rather than producing the same output.
+
+When the document you're writing is a skill, read [`SKILL-MECHANICS.md`](SKILL-MECHANICS.md) for frontmatter, invocation choice, and router skills.
+
+## Context pointers
+
+A **context pointer** is a reference held in the agent's context that names some out-of-context material and encodes the condition for reaching it. A skill's description is one; a line in `AGENTS.md` naming a doc is the same object. The pointer's _wording_, not its target, decides when the agent reaches the material, and how reliably. A must-have target behind a weakly worded pointer is a variance bug: sharpen the wording first, and inline the material only if sharpening fails.
+
+A pointer does two jobs: state what the material is, and list the **branches** that should trigger reaching it (a branch is a distinct case the document handles, so different runs take different paths through it). Every word of an always-loaded pointer costs on every turn, so it earns even harder pruning than the body:
+
+- **Front-load the leading word**: the pointer is where it does its triggering work.
+- **One trigger per branch.** Synonyms that rename a single branch are one branch written twice; collapse them and keep only genuinely distinct branches.
+- **Cut identity the body already carries.**
+
+## The two loads
+
+Every document and pointer you add spends one of two budgets:
+
+- **Context load** is the cost of always-loaded material on the agent's window: an `AGENTS.md` line, a skill description, anything sitting in context every turn, spending tokens and attention whether or not it fires.
+- **Cognitive load** is the cost on the human: which documents exist and when to reach for each. The human is the index. Not a cost to minimise: it is the price of human agency; spend it where human judgement matters, remove it where it does not.
+
+Material reached only through a pointer escapes context load at the price of the pointer's own line; material with no pointer at all rides entirely on cognitive load.
+
+## Information hierarchy
+
+A document is built from two content types: **steps** (the ordered actions the agent performs) and **reference** (definitions, rules, facts consulted on demand). The two mix freely: all steps (a recipe), all reference (a review's rules, this skill), or both. The core decision is where each piece sits on the **information hierarchy**, a ladder ranked by how immediately the agent needs the material:
+
+1. **In-file step** is the primary tier: what the agent does, in order.
+2. **In-file reference** is consulted on demand. Often a legitimately flat peer-set (every rule of a review on one rung), which is a fine arrangement, not a smell.
+3. **Disclosed reference** is pushed out into a separate file, reached by a context pointer, loaded only when the pointer fires. Spans a sibling file in the same folder through fully external reference that lives anywhere and any document can point at.
+
+Push too little down and the top bloats; push too much and you hide material the agent actually needs. That tension is the whole decision.
+
+**Progressive disclosure** is the move down the ladder (out of the main file and behind a pointer) so the top stays legible. Not primarily a token optimisation: it is how the hierarchy is protected. Branching is the cleanest disclosure test: inline what every branch needs, and push behind a pointer what only some branches reach. When a document has steps, in-file reference that should be disclosed buries them and turns attending to them into a coin-flip: a variance lever, not just a legibility one.
+
+**Co-location** is the within-file companion: where the ladder decides _how far down_ a piece sits, co-location decides _what sits beside it_ once there. Keep a concept's definition, rules, and caveats under one heading rather than scattered, so reading one part brings its neighbours with it. The test: the document should read like documentation written for the agent. Grouped material reads that way; scattered material does not. (Distinct from duplication: that repeats one meaning in two places; scattering fragments one meaning across many.)
+
+**Sprawl** is the failure mode here: a document simply too long, even when every line is live and unique. Attention thins across the excess, and every extra line is one more to keep relevant. The cure is the ladder: disclose reference behind pointers, and split by branch or sequence so each path carries only what it needs.
+
+## Steps and completion criteria
+
+Every step ends on a **completion criterion**, the condition that tells the agent the work is done. Two properties make it a lever:
+
+- **Clarity**: can the agent tell done from not-done? A vague bound ("understanding reached") invites **premature completion**: ending the step before it is genuinely done, attention slipping to _being done_. The visible steps still ahead (the **post-completion steps**) supply the pull; the criterion's clarity is the resistance. Defend in order: **sharpen the bound first** (local and cheap); only if it is irreducibly fuzzy _and_ you observe the rush, hide the later steps by splitting the sequence. Hiding only works across a real context boundary (a hand-off or a subagent dispatch; an inline call leaves the later steps in context and clears nothing).
+- **Demand**: how much it requires. "Every modified model accounted for" forces thorough work where "produce a change list" does not. Demand drives **legwork** (the digging the agent does within the work, latent in the wording rather than written as its own step), and it is not step-bound: "every rule applied" binds a body of flat reference just as "every step done" binds a sequence, which is how an all-reference document still carries an exhaustiveness bar.
+
+The strongest criteria are both checkable and exhaustive.
+
+## When to split
+
+Splitting one document into two spends one of the two loads, so split only when the cut earns it:
+
+- **By sequence**: split a run of steps where the post-completion steps tempt the agent to rush the one in front of it. Keeping them out of view drives more legwork on the current task. Beware the reverse: merging sequences exposes each step's later steps to what follows, inviting premature completion.
+- **By invocation**, skill-specific: see [`SKILL-MECHANICS.md`](SKILL-MECHANICS.md).
+
+## Leading words
+
+A **leading word** is a compact concept already living in the model's pretraining that the agent thinks with while running the document (_lesson_, _fog of war_, _tracer bullets_). Repeated as a token, never as a sentence, it accumulates a distributed definition and anchors a whole region of behaviour in the fewest tokens, by recruiting priors the model already holds. Coining your own works if you define it clearly, but a made-up word recruits no priors: you pay in definition tokens what a pretrained word gives free; reach for an existing word first.
+
+It anchors twice. In the body, _execution_: the agent reaches for the same behaviour every time the word appears, and inside flat reference it focuses attention on a class of thing to look for. In a pointer, _invocation_: when the same word lives in your prompts, your docs, and your codebase, the agent links that shared language to the material and reaches it more reliably.
+
+Hunt for opportunities to refactor with leading words. A triad spelled out at three sites, a pointer spending a sentence to gesture at one idea. Each is a passage begging to collapse into a single token:
+
+- "fast, deterministic, low-overhead" → _tight_ (a _tight_ loop).
+- "a loop you believe in" → _red_, turning a fuzzy gate into a binary observable state (the loop goes _red_ on the bug, or it doesn't).
+
+You win twice: fewer tokens, and a sharper hook for the agent to hang its thinking on. Assume every document is carrying restatements that leading words retire. Go find them.
+
+**Negation** is the failure mode beside this lever: steering by prohibition drags the forbidden behaviour into context and makes it _more_ available, not less. _Don't think of an elephant_, and the elephant is all there is; the negation is a weak modifier the strongly-activated concept overruns, so the ban half-reads as an instruction to do the thing. Prompt the **positive**: state the target behaviour ("write one-line comments") so the banned one is never spoken. A prohibition earns its place only as a hard guardrail you cannot phrase positively; even then, pair it with the positive target so attention lands on what to do.
+
+## Pruning
+
+- Keep each meaning in a **single source of truth**: one authoritative place, so changing the behaviour is a one-place edit. **Duplication** (the same meaning in more than one place) costs maintenance and tokens, and inflates a meaning's prominence on the ladder past its real rank. (The accidental inverse of a leading word, which repeats a token on purpose, never the meaning.)
+- The **environment** is a source of truth too (`package.json` scripts, config files, the directory layout, `--help` output), and a document that restates it is a **cache**: a copy of a lookup, earning its load only when the lookup is expensive. Cache what the agent cannot find by looking: the unwritten convention, the reason behind a choice, the gotcha no config confesses. Leave the one-file, one-command lookups to the environment, where they cannot go stale.
+- Check every line for **relevance**: does it still bear on what the document does? A line loses relevance by never bearing on the task (mere exposition, or a branch that should be disclosed) or by going stale as the behaviour or world it describes changes. Shorter documents are easier to keep relevant. Without a pruning discipline the default fate is **sediment**: stale layers that settle because adding feels safe and removing feels risky, until you must core down through them to find what is still live.
+- Hunt **no-ops** sentence by sentence: an instruction the model already obeys by default pays load to say nothing. The test (does it change behaviour versus the default?) is model-relative, not reader-relative: two people disagreeing about a no-op disagree about the default, and settle it by running the document, not by debate. When a sentence fails, delete the whole sentence rather than trim words from it. The test also grades leading words: a word too weak to beat the default (_be thorough_ when the agent is already thorough-ish) is a no-op, and the fix is a stronger word (_relentless_), not a different technique.

src/skills/craft-skill/GLOSSARY.md

@@ -1,161 +0,0 @@
-# Glossary — Skill Craft
-
-This glossary defines the checks for [`craft-skill`](SKILL.md).
-Each term is a lever on **predictability**.
-
-## Predictability
-
-The degree to which a skill makes the agent follow the same process on every run.
-Predictability does not require identical output.
-A brainstorming skill can predictably produce different ideas.
-
-## Invocation
-
-How the agent or operator reaches a skill.
-Invocation controls who pays the load for discovery.
-
-### Model-invoked
-
-A skill that the harness exposes to the agent's available-skills list.
-The agent can load it without an explicit operator command.
-Use model invocation when the agent must discover the skill on its own.
-
-### User-invoked
-
-A skill that the operator runs deliberately through the harness invocation mechanism.
-Use user invocation when human judgment decides when the skill applies.
-For Pi, keep the required description and add `disable-model-invocation: true`.
-
-### Description
-
-Required metadata that states what the skill does and when to use it.
-For model-invoked skills, the description carries the trigger condition.
-For user-invoked skills, the description remains valid metadata.
-
-### Context pointer
-
-A short reference that tells the agent when to load out-of-context material.
-The pointer's wording controls whether the agent reaches the target.
-Fix weak wording before moving the target inline.
-The description is the top-level context pointer for a model-invoked skill.
-A link to a sibling reference file is a lower-level context pointer.
-
-### Context load
-
-The permanent attention cost of model-invoked skill descriptions.
-More discoverable skills create more context load.
-Use context load to decide whether a skill needs model invocation.
-
-### Cognitive load
-
-The memory cost that user-invoked skills place on the operator.
-More user-invoked skills make the operator remember more names and uses.
-Use cognitive load to decide whether a skill needs discoverability.
-
-## Information hierarchy
-
-How a skill ranks information by the moment the agent needs it.
-Keep the active procedure above supporting reference.
-Move branch-specific or large reference behind context pointers.
-
-### Steps
-
-Ordered actions that the agent performs.
-Put steps in `SKILL.md` when the skill has an active procedure.
-End each step with a completion criterion.
-
-### Reference
-
-Material that the agent consults while following steps.
-Keep short reference near the step that needs it.
-Move large or branch-specific reference to sibling files.
-
-### External reference
-
-Authoritative material that lives outside the skill directory.
-Point to external reference instead of duplicating project policy.
-Use it when another file owns the meaning.
-
-### Progressive disclosure
-
-Moving reference out of `SKILL.md` and behind a precise context pointer.
-Use progressive disclosure to protect the skill's active procedure.
-Keep must-read material inline when a pointer would miss too often.
-
-### Co-location
-
-Keeping a concept's rules, caveats, and examples near one another.
-Use co-location so one read gives the agent the complete local meaning.
-Do not scatter one concept across unrelated sections.
-
-## Steering
-
-How a skill shapes the agent's runtime behavior.
-Steering turns a document into a repeatable process.
-
-### Branch
-
-A distinct path through a skill.
-A branch handles a case that does not run the same steps as other cases.
-Do not call ordinary checklist items branches.
-
-### Leading word
-
-A compact term that anchors a repeated behavior.
-Prefer words already used by prompts, docs, or code.
-Repeat the word, not the full explanation.
-
-### Completion criterion
-
-The condition that tells the agent a step is complete.
-A strong completion criterion is checkable and demanding.
-Use it to prevent early movement to the next step.
-
-### Legwork
-
-The work the agent performs inside a step.
-Strong completion criteria increase legwork.
-Weak criteria let the agent stop after surface-level work.
-
-### Post-completion steps
-
-The steps visible after the current step.
-Visible later steps can pull the agent forward too early.
-Hide later steps only when a real context boundary exists.
-
-## Failure modes
-
-Use these findings during audit and verification.
-Quote the offending passage when a finding applies.
-
-### Premature completion
-
-A step ends before its necessary work is complete.
-This usually comes from a vague completion criterion.
-Fix the local criterion before splitting the skill.
-
-### Duplication
-
-One meaning has more than one authoritative expression.
-Duplication increases maintenance cost and false emphasis.
-Replace duplicates with one source and pointers to it.
-
-### Sediment
-
-Old content remains after the skill's behavior changes.
-Sediment makes the agent attend to stale instructions.
-Remove lines that no longer affect the skill's current behavior.
-
-### Sprawl
-
-`SKILL.md` contains too much live material.
-Sprawl hides the active procedure among supporting details.
-Move reference down the information hierarchy.
-
-### No-op
-
-An instruction does not change the agent's default behavior.
-No-op instructions spend attention without adding control.
-Judge no-op instructions against behavior, not style.
-Delete a no-op sentence instead of shortening it.
-Cut no-op sentences unless they anchor a target-project convention.

src/skills/craft-skill/SKILL.md

@@ -1,116 +0,0 @@
----
-name: craft-skill
-description: Draft or audit an Agent Skill for structural predictability. Run deliberately as /skill:craft-skill when creating or improving a skill.
-disable-model-invocation: true
----
-
-# craft-skill
-
-Draft or audit an Agent Skill.
-Judge the skill against one bar: **predictability**.
-Predictability means the agent follows the same process on every run.
-
-Read [`GLOSSARY.md`](GLOSSARY.md) before applying glossary terms.
-Use this skill for skill structure, not general prose review.
-Apply project writing rules when the target project defines them.
-
-## Choose the branch
-
-Use **Draft** when creating a new skill source.
-Use **Audit** when reviewing an existing skill in its authoritative source.
-
-Finish when one branch matches the user's requested change.
-
-## Draft
-
-1. Resolve the destination.
-   Read the target project's instructions and skill-source conventions.
-   Prefer an existing source directory over a new convention.
-   Trace symlinks and generated placements to their source.
-   Ask the user when multiple writable sources remain plausible.
-
-   Finish when the destination is writable and authoritative.
-
-2. Create the skill structure.
-   Create a directory whose name matches the skill name.
-   Add `SKILL.md` with valid Agent Skills frontmatter.
-   Add `disable-model-invocation: true` only for Pi user-invoked skills.
-   Add sibling files only when they protect the **information hierarchy**.
-
-   Finish when the file layout matches the target harness and project.
-
-3. Write the skill body.
-   Put active **steps** in `SKILL.md`.
-   Keep shared **reference** near the steps that need it.
-   Move large or branch-specific reference behind precise **context pointers**.
-   Front-load the **leading word** in descriptions and other context pointers.
-   Write pointer text for the condition that loads the target.
-   Give each step a checkable **completion criterion**.
-   Cut sentences that do not change agent behavior.
-   Use relative paths for all skill-local references.
-
-   Finish when each branch has ordered steps, resolved references, and checkable completion criteria.
-
-## Audit
-
-1. Locate the authoritative source.
-   Read the target project's instructions and skill-source conventions.
-   Inspect enabled-skill directories only to find the source.
-   Do not edit symlinks, generated files, installed packages, or vendored copies.
-   Ask the user when multiple sources claim the same skill.
-
-   Finish when you have identified the authoritative source.
-
-2. Read the skill surface.
-   Read `SKILL.md` and every disclosed file.
-   Read project documents that the skill treats as authoritative.
-   Record which file owns each meaning before judging duplication.
-
-   Finish when all referenced behavior is visible.
-
-3. Apply the checklist.
-   Quote each passage that fails a glossary check.
-   Check **invocation** against the intended harness.
-   Check **context pointers** for weak trigger wording.
-   Check **information hierarchy** for buried steps or disclosed required material.
-   Check **completion criteria** for vague or weak stopping points.
-   Check **leading words** for repeated explanations they can replace.
-   Check **pruning** for sentences that explain default model behavior.
-   Check **duplication**, **sediment**, **sprawl**, and **no-op** instructions.
-
-   Finish when every finding has quoted evidence or no finding remains.
-
-4. Report or fix findings.
-   Report findings by default.
-   Fix findings only when the user requests edits.
-   Preserve behavior that passes the checklist.
-   Change only structural skill problems unless the user expands the scope.
-
-   Finish when you have reported each finding or fixed each requested finding.
-
-## Verify
-
-1. Run deterministic checks.
-   Verify required frontmatter exists.
-   Verify the name is portable and matches the directory.
-   Verify the description is specific and under 1024 characters.
-   Verify every skill-local reference resolves with a relative path.
-   Verify context pointers name when to load referenced material.
-   Verify disclosed files exist.
-   Verify each retained sentence changes behavior or records target-project policy.
-   Verify no stale harness-specific assumption remains.
-
-   Finish when each check passes or the final report names checks that could not run.
-
-2. Re-read against the glossary.
-   Re-check **invocation**, **information hierarchy**, and **completion criteria**.
-   Re-check **duplication**, **sediment**, **sprawl**, and **no-op** instructions.
-   Fix any new structural finding.
-
-   Finish when no unresolved structural finding remains.
-
-3. Report the result.
-   List every changed path.
-   State whether you enabled the skill or only authored it.
-
-   Finish when the user can review the exact files changed.

src/skills/grill/SKILL.md

@@ -1,88 +1,41 @@
 ---
 name: grill
-description: Stress-test a plan, decision, or design through a one-question-at-a-time interview. Use when the user asks to grill, stress-test, pressure-test, challenge, interrogate, or poke holes in a plan.
+description: Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.
 ---
 
-# grill
+Interview the user relentlessly until you reach a shared understanding. Map this as a **design tree**: every decision branches into the decisions that hang off it.
 
-Interview the user until the session has no open or blocked questions.
+Maintain the design tree in a new Markdown file. Use the file as the source of truth.
-Open questions define the ending condition for the grill session.
-Blocked questions track known decisions with unsettled prerequisites.
 
-Do not begin implementation during or after grill without an explicit user instruction.
+Work the tree one question at a time. The **frontier** is every decision whose prerequisites are already settled: the questions you can ask now without guessing at answers you have not heard yet. Record the next frontier question in the file, then ask it and wait for the user's answer.
-A completed grill summary is not an instruction to build.
 
-## Run the interview
+Format a question like so:
 
-1. Prepare the grill.
+```
-   Read the conversation, files, and docs that can settle facts or reveal constraints.
+❓ **Q1** - **<question title>**: <question body, might be multiple paragraphs, including multiple choices>
-   Resolve factual questions yourself when tools can answer them.
-   Do not ask the user for facts you can find.
-   Build the required-question set from the known plan or idea.
-   Put ready questions under `Open questions:`.
-   Put known questions with unsettled prerequisites under `Blocked questions:`.
 
-   Finish when both sections contain every known required question.
+**Options**
 
-2. Start the interview.
+1. <option>
-   Show `Open questions:` and `Blocked questions:` in short form.
+...
-   Include titles only, not full question bodies.
+<n>. **<recommended option>** — recommended
-   Use Markdown list items in both sections.
-   Mark each initial item with `- [ ]`.
-   Ask exactly one numbered question in the same turn.
-   The question title must match an item under `Open questions:`.
 
-   Finish when the interview has started with exactly one numbered question.
+---
 
-3. Process the answer.
+❓ **Q2** - **<question title>**: <question body, might be multiple paragraphs, including multiple choices>
-   Account for the user's answer before asking another question.
-   Resolve, update, remove, unblock, or add questions as the answer requires.
-   Use `- [ ]` for unchanged unresolved questions.
-   Use `- [+]` for questions added by the latest answer.
-   Use `- [u]` for questions updated by the latest answer.
-   Use `- [x]` for questions resolved by the latest answer.
-   Use `- [-]` for questions removed by the latest answer.
-   Include a cause note for each `- [+]`, `- [u]`, and `- [-]` change.
-   Indent cause notes by two spaces under the changed question.
-   Show `- [+]`, `- [u]`, `- [x]`, and `- [-]` for one turn only.
-   After that turn, keep an updated question visible as `- [ ]`.
-   After that turn, remove resolved and removed questions from the visible list.
-   Move unblocked questions from `Blocked questions:` to `Open questions:` with `- [u]`.
-   Do not use `- [+]` for unblocked questions because they are not new.
-   Add questions only when the latest answer directly causes them.
-   When the user extends scope instead of answering, account for the new question before pursuing it.
-   Do not add optional follow-up questions to either section.
-   Put optional follow-up questions in the final summary instead.
-   Resolve newly answerable facts yourself.
 
-   Finish when every list change has an explicit marker and required cause note.
+**Options**
 
-4. Ask the next question.
+1. <option>
-   Show `Open questions:` and `Blocked questions:` on every active grill turn.
+...
-   Keep `- [ ]` items visible every turn.
+<n>. **<recommended option>** — recommended
-   Preserve agent judgment when the user reframes or invalidates a question.
+```
-   Show the resulting list change instead of following a rigid reframe algorithm.
-   If the user catches incorrect accounting, briefly explain the error.
-   Then show the corrected question sections.
-   Ask exactly one numbered question when `Open questions:` contains an unresolved item.
-   The question title must match an item currently shown under `Open questions:`.
-   Give concrete options with trade-offs.
-   Label options with `A.`, `B.`, `C.`, and later letters.
-   Use the same labels in the recommendation.
-   Give a recommendation and rationale.
-   Probe vague words, overloaded terms, and hidden assumptions.
-   Use concrete scenarios to test concept boundaries and edge cases.
-   When code contradicts the user's claim, quote both the claim and the code behavior.
-   If the turn includes a correction, file edit, apology, or explanation, still end with the next numbered question.
 
-   Finish when the turn ends with the next numbered question.
+List every viable answer in an ordered list and mark your recommended answer.
+Word each question so "yes" accepts your recommended answer.
 
-5. Close or pause.
+Each user answer reshapes the tree: settled decisions push the frontier outward and unblock questions that depended on them. Recompute the frontier in the file and ask the next question. A question whose answer depends on another open question belongs later, not next.
-   Close when `Open questions:` and `Blocked questions:` are empty.
-   Output a concise summary of every decision reached.
-   Include optional follow-up questions that were not added to either section.
-   Do not act on the summary unless the user gives a new implementation instruction.
-   If the user changes tasks, state that grill is paused.
-   End the turn with the final decision summary or the pause statement.
 
-   Finish when the summary is delivered or grill is explicitly paused.
+Finding _facts_ is your job, never the user's. When a frontier question needs a fact from the environment (filesystem, tools, etc.), find it with available tools; don't ask the user for anything you could look up yourself. The _decisions_ are the user's: put each to them and wait.
+
+The session is done when the frontier is empty: every branch of the design tree visited, nothing left silently assumed. Present every settled decision from the design tree. Do not act on it until the user confirms you have reached a shared understanding.

src/skills/retrospect/SKILL.md

@@ -0,0 +1,44 @@
+---
+name: retrospect
+description: "Conduct a retrospective on a coding session."
+disable-model-invocation: true
+---
+
+The user has asked for a **retrospective**. You are suggesting improvements to the coding agent's **environment** to improve future runs.
+
+## Steps
+
+1. Read [`../agent-writing-guide/SKILL.md`](../agent-writing-guide/SKILL.md).
+
+2. Read the primary sources for the session the user specifies. This may mean searching through session logs on this machine. If the user doesn't specify a session, default to the current one.
+
+3. Look for candidates for improvement in these categories.
+
+- **Navigation**: how easy was it for the agent to find the right files? Are there hidden dependencies between files? Would a **navigation pointer** make it easier? _Use when_ the session took a long time to find a piece of information.
+- **Automated checks**: are there automated checks that could catch errors the agent made? Linting, typing, tests, filesystem linters? Read the repo's own check command first (its `package.json`/build-tool `lint`/`check` scripts, its CI workflow), so a check that already exists but sits unwired or silently broken is the finding, not a reinvention. A repo with no **guardrail** (no pre-commit hook and no CI job running its lint/typecheck/test command) is itself a finding: an un-linted repo is a standing missed opportunity, not a neutral default. _Use when_ the agent made a mistake an automated check could have caught, or the repo has no guardrail at all.
+- **Coding standards**: should the **reviewer agent** be given a new rule to enforce? Should an existing rule be removed or clarified? Classify the violation first: a **mechanical** one (a fixed syntactic pattern, a banned API, an import shape, a file-location rule) gets a deterministic check, full stop: a custom rule in the repo's own linter, a new pre-commit hook, or a new CI job, whichever the repo's language and existing guardrail make cheapest. Default to building the check over writing the rule. Reserve **review standards** for genuine **judgement calls** (cross-file consistency, "matches the surrounding style," anything no guardrail could ever substitute for). Before proposing a rule, identify the repository source that its review workflow loads. Keep the rule in that established source. _Use when_ the reviewer agent failed to catch a mistake.
+- **Global AGENTS.md**: are there any steering instructions that should be moved to review standards (or automated checks) instead? _Use when_ the AGENTS.md file is particularly large - in the repo OR the user's global scope.
+- **Tool economy**: did the agent make expensive tool calls that could be streamlined? Is there any custom tooling (CLI's, MCP's) that is particularly token-inefficient? _Use when_ the agent made an expensive tool call.
+- **No-ops**: look for instructions in steering files that don't modify the agent's behavior. _Use when_ the steering files are large and unwieldy.
+- **Information access**: look for opportunities to increase the agent's access to information. Teeing dev server logs, readonly access to third-party services. _Use when_ a crucial piece of information was not available to the agent.
+
+4. Present these candidates to the user, in order of severity.
+
+## Reference
+
+### Implementation vs Review
+
+Remember that all work goes through two stages: implementation and review. The implementation agent has the most **context pressure**. They are responsible for exploration, writing code, and debugging failures.
+
+The review agent has the least context pressure - it receives a diff, so no exploration needed. It often does not need to write code or debug.
+
+This means that the review agent should be responsible for imposing coding standards, not the implementation agent.
+
+### Files
+
+You have access to several files in the repo:
+
+- `AGENTS.md`: these files are pushed to the context window of any agent working in this repo. They should be used incredibly sparingly, usually only for **navigation pointers** to other files.
+- **Review standards**: these are the documented rules that the repository's review workflow loads. They can live in a standards file, a contribution guide, or scoped rule files. Find the existing source before proposing an addition.
+- Docs: use docs as references files, pointed to by other files. Look for existing docs before writing new ones.
+- Skills: use skills for user-invoked commands or model-discoverable guidance. Follow the advice in the `agent-writing-guide` skill.

src/skills/style/nix/rules/dotfiles.md

@@ -0,0 +1,43 @@
+## Register window icons
+
+### Rule
+
+Each graphical application has a `registry.window-icon` entry with a window-class match and rendered icon.
+
+### Rationale
+
+Hyprland uses this registry to render workspace icons for open application windows.
+
+Run this command to list window classes from the running compositor:
+
+```bash
+hyprctl clients -j | jq -r '.[].class' | sort -u
+```
+
+## Separate guest wrappers from service modules
+
+### Rule
+
+Each guest module wraps one service module.
+The guest enables its service through `guests.<name>.interior`.
+Service modules own service configuration.
+Guest modules configure bind mounts, secret delivery, resource limits, and registry entries.
+Host configuration supplies deployment-specific paths, volumes, addresses, and routes.
+
+### Rationale
+
+Service modules support a service without guest infrastructure.
+Guest modules adapt a service module to a guest.
+Host configuration provides deployment-specific values.
+
+## Decrypt guest secrets on the host
+
+### Rule
+
+A host user owns each guest secret before SOPS decrypts it.
+Each guest service group accesses its secrets through a read-only bind mount.
+
+### Rationale
+
+Guest service accounts do not exist on the host, where SOPS creates secrets.
+The shared service group preserves read access across the guest boundary.

src/skills/style/nix/rules/general.md

@@ -22,6 +22,17 @@ Do not call `lib.removeSuffix ".nix"` outside the utility definition.
 
 The utility centralizes Nix file-name normalization.
 
+## Qualify library functions
+
+### Rule
+
+Keep library functions qualified as `lib.<function>`.
+
+### Rationale
+
+Qualification makes library ownership explicit.
+It prevents local namespace collisions.
+
 ## Use executable helpers for package commands
 
 ### Rule

src/skills/to-document/DOCUMENT-FORMAT.md

@@ -1,62 +0,0 @@
-# Document Format
-
-## Template
-
-```md
-## Problem Statement
-
-The problem that the user is facing, from the user's perspective.
-
-## Solution
-
-The solution to the problem, from the user's perspective.
-
-## User Stories
-
-An extensive, numbered list of user stories, covering all aspects of the feature. Each user story should be in the format of:
-
-1. As an <actor>, I want a <feature>, so that <benefit>
-
-<user-story-example>
-1. As a mobile bank customer, I want to see balance on my accounts, so that I can make better informed decisions about my spending
-</user-story-example>
-
-## Implementation Decisions
-
-A list of implementation decisions that were made. This can include:
-
-- The modules that will be built/modified
-- The interfaces of those modules that will be modified
-- Technical clarifications from the developer
-- Architectural decisions
-- Schema changes
-- API contracts
-- Specific interactions
-
-## Testing Decisions
-
-A list of testing decisions that were made. Include:
-
-- A description of what makes a good test (only test external behavior, not implementation details)
-- Which modules will be tested
-- Prior art for the tests, that is, similar tests in the codebase
-
-## Out of Scope
-
-A description of the things that are out of scope for this spec.
-
-## Further Notes
-
-Any further notes about the feature.
-```
-
-## Rules
-
-- **Do not include specific file paths or code snippets.**
-  They can become outdated quickly.
-  Exception: include a prototype snippet when it records a decision more precisely than prose.
-  Examples include a state machine, reducer, schema, or type shape.
-  Inline the snippet in the relevant implementation decision.
-  Note briefly that the snippet came from a prototype.
-  Trim the snippet to the decision-rich parts.
-  Do not include a working demo.

src/skills/to-document/SKILL.md

@@ -1,64 +0,0 @@
----
-name: to-document
-description: Draft a requested summary, specification, plan, decision record, or similar document from conversation and repository context.
----
-
-Do not start with an interview when the conversation gives a usable subject.
-Ask before drafting only when the conversation lacks a usable subject.
-
-Do not put draft contents in the chat response.
-Always persist the draft to a file.
-
-Do not implement the draft contents unless the operator gives a separate instruction.
-Treat a completed draft as documentation, not an instruction to build.
-
-## Process
-
-1. Identify the subject.
-   Use the conversation as the primary source.
-   If the conversation lacks a usable subject, ask what the document is about.
-   Do not ask clarifying questions before drafting when a usable subject exists.
-
-   Finish when you have a subject or a question that requests the missing subject.
-
-2. Gather repository facts that the draft needs.
-   Inspect files when repository details matter.
-   Do not invent modules, paths, commands, flows, contracts, or conventions.
-   Distinguish confirmed facts from assumptions.
-
-   Finish when each repository-specific claim has support or an explicit assumption marker.
-
-3. Choose the document shape.
-   Infer the document type and title from context.
-   Load [DOCUMENT-FORMAT.md](DOCUMENT-FORMAT.md) when the operator requests no specific format.
-   Adapt the structure only when the requested document type needs different sections.
-   Preserve useful sections for decisions, testing, scope, and notes.
-
-   Finish when the draft has a title and section plan that fit the requested document.
-
-4. Write the draft.
-   Capture known context accurately.
-   Mark assumptions and gaps explicitly.
-   Include decisions that the conversation settled.
-   Include open questions only as gaps, not as an interview.
-   Keep the document focused on the operator's problem and intended use.
-
-   Finish when the draft contains the relevant facts, decisions, assumptions, gaps, and scope.
-
-5. Persist the draft.
-   Use `docs/00 INBOX/ai` as the default destination.
-   Use the operator's requested path when one exists.
-   Derive a kebab-case slug from the inferred title.
-   Create the destination directory when it does not exist.
-   If the target file exists, summarize the expected change.
-   Confirm before overwriting an existing file.
-   Do not print the draft in chat instead of writing it.
-
-   Finish when the draft is written or overwrite confirmation is pending.
-
-6. Finish with the result.
-   Report the written path.
-   Do not include the draft contents in the response.
-   Do not continue into implementation without an explicit instruction.
-
-   Finish when the operator can find the written draft.