Commit Graph
2520 Commits
Author SHA1 Message Date
ryan4yin | 二花 68ea04f173 fix(docs): correct stale and mismatched documentation (#338) 2026-10-02 10:21:55 +08:00
ryan4yin | 二花 b2ec56db66 docs(skills): add repo-scoped task skills and give each fact one home (#337) 2026-10-02 10:13:03 +08:00
Ryan Yin 1f8db1fccf fix(catppuccin): drop the ignored user-level cache entry
Catppuccin's home-manager module writes extra-substituters/extra-trusted-public-keys
for catppuccin.cachix.org into the user-level ~/.config/nix/nix.conf. Since the
normal user is no longer a trusted-user (modules/base/nix.nix), Nix ignores those
client-specified settings and warns on every command. The ports build locally, so
turn the user-level cache off instead.
2026-10-01 21:25:55 +08:00
ryan4yin | 二花 739cd171ba feat(neovim): replace nixvim with programs.neovim (#336)
nixvim has been disabled for a long time (programs.nixvim.enable = false), so
drop the flake input (and prune flake.lock). Port the settings it carried to
Home Manager's programs.neovim: opts and per-plugin setup become an initLua
block plus a plugin list. The catppuccin module already provides the
colorscheme, vi stays unmapped, and the plain system neovim package is
untouched so sudo/nvim --clean keep using it.
2026-10-01 21:19:18 +08:00
ryan4yin | 二花 4909f6357b security: scope privileges and stop a world-readable secret copy (#335) 2026-10-01 21:18:34 +08:00
ryan4yin | 二花 d690f45183 feat: add CanoKey hardware security key support (#334) 2026-10-01 20:50:13 +08:00
Ryan Yin 14db49660b chore(preservation): drop persist entries for removed apps 2026-10-01 20:10:08 +08:00
Ryan Yin a6553b25a5 chore: drop archived terraformer and unused netbird 2026-10-01 20:10:07 +08:00
ryan4yin | 二花 ab2fd0fee9 feat: add plocate filesystem search and cap podman log size (#333) 2026-10-01 20:00:53 +08:00
ryan4yin | 二花 8f9fdda2bb refactor: trim unused virtualisation config and packages (#332) 2026-10-01 19:42:20 +08:00
Ryan Yin 1546e54bb2 chore(noctalia): move OSD and notifications to the bottom-right 2026-10-01 14:09:16 +08:00
Ryan Yin 872901007e docs(agents): use terraform plan, drop nonexistent helm diff 2026-10-01 11:34:34 +08:00
Ryan Yin ea73c716e8 flake.lock: Update
Flake lock file updates:

• Updated input 'mysecrets':
    'git+ssh://git@github.com/ryan4yin/nix-secrets.git?ref=refs/heads/main&rev=6daee5c0fa300bfcd13d702ae1d69aeb0a4c281d&shallow=1' (2026-09-28)
  → 'git+ssh://git@github.com/ryan4yin/nix-secrets.git?ref=refs/heads/main&rev=decf57110a6ecbd45b3ab7ac5ec7db393f17d3a7&shallow=1' (2026-10-01)
2026-10-01 10:53:08 +08:00
Ryan Yin 9885283edf docs(agents): clarify local language selection and prefer Bun 2026-10-01 10:23:35 +08:00
Ryan Yin 01ed0aa52f chore(agents): upgrade to opencode v2 2026-10-01 09:39:05 +08:00
Ryan Yin 7f6759219b docs(agents): align evals with external instruction boundaries 2026-09-30 21:17:19 +08:00
Ryan Yin 91cc9c0cf4 docs(agents): fix authorization and skill trust wording
Allow clearly-implied task authorization, trust skills listed in the runtime's configured catalog, and restore the ban on following external-material instructions. Sync the affected evaluation scenarios.
2026-09-30 21:09:34 +08:00
ryan4yin | 二花 111929023d docs(agents): clarify tool selection and authorization boundaries (#331) 2026-09-30 20:51:42 +08:00
Ryan Yin bae98ed1c8 docs(agents): allow TypeScript for complex shell commands 2026-09-30 18:38:08 +08:00
Ryan Yin e10f0d6794 flake.lock: Update
Flake lock file updates:

• Updated input 'llm-agents':
    'github:numtide/llm-agents.nix/03aa73a454d7468e0a3d44faee511fa6ae7c7ded?narHash=sha256-DjGLaiqnHlHQ9d8EfKhklMlCRoEC97FkI9XRgznopL8%3D' (2026-09-26)
  → 'github:numtide/llm-agents.nix/af40d966859ec4075ecc172dbb39e53f474dc5d9?narHash=sha256-I4O5GSbZgmvzV7dj8fZN8furV2UdEpahh4rvsZOfs/Q%3D' (2026-09-30)
• Updated input 'llm-agents/nixpkgs':
    'github:NixOS/nixpkgs/d54020a6ac3211e9f4201631bdf67678818c0cdf?narHash=sha256-StRyS/lFHtIw/UgJk%2Bxo%2BHA66mtoWeDUbkJdvf2Byn4%3D' (2026-09-26)
  → 'github:NixOS/nixpkgs/f45c6f04c2f013f004bf94e284e95d72898d9393?narHash=sha256-6k3n7X2UsZ6tAlWTz99RimiZab%2B17tMUrw9sRUcELiA%3D' (2026-09-28)
2026-09-30 16:12:45 +08:00
ryan4yin | 二花 9ab5eca60c feat(monitoring): GPU-aware btop variants and GPU metrics (#330) 2026-09-30 00:16:50 +08:00
ryan4yin | 二花 523f4940a7 Merge pull request #329 from ryan4yin/youko-shushou-ssd-swap
hosts: swap youko/shushou roles to fix Jellyfin HDR transcoding
2026-09-29 22:02:11 +08:00
Ryan Yin c7d8db2bf3 fix(k8s): rebalance k3s-test-1 VMs after the youko/shushou swap 2026-09-29 21:46:09 +08:00
Ryan Yin 2e5f539a47 fix(hosts): move UPS/nut metrics to youko after the swap 2026-09-29 21:36:47 +08:00
Ryan Yin 2be0337979 docs(hosts): record the youko/shushou swap and its rationale 2026-09-29 21:36:47 +08:00
ryan4yin | 二花 8ef52e2cf7 fix(youko): add ROCm OpenCL so Jellyfin can tone-map HDR (#328)
* fix(youko): add ROCm OpenCL so Jellyfin can tone-map HDR

Jellyfin's HDR/DoVi transcodes use ffmpeg's `tonemap_opencl`, but only
Mesa was installed via `hardware.graphics.enable`, so no OpenCL ICD was
registered. ffmpeg then failed to create an OpenCL device ("Failed to
get number of OpenCL platforms: -1001") and aborted with exit code 237,
leaving HDR titles unplayable.

Add `rocmPackages.clr.icd` to `hardware.graphics.extraPackages`; it
ships the ICD for the Barceló iGPU (reported as `gfx90c`) under
/run/opengl-driver/etc/OpenCL/vendors, where `ocl-icd` (built with
`--enable-custom-vendordir`) finds it.

The Barceló iGPU only manages ~0.8x realtime for a 4K HDR -> 1080p
tone-mapped transcode, so playback can still buffer; this only makes
such titles playable at all.

* fix(youko): allow /dev/kfd so Jellyfin OpenCL can reach the GPU

The jellyfin module's `DeviceAllow` whitelist only permits the VA-API
render node (`/dev/dri/renderD128`). ROCm's OpenCL runtime opens
`/dev/kfd`, so even with the ICD installed OpenCL device creation failed
("No matching devices found") and HDR transcodes still aborted with exit
code 237.

Add `/dev/kfd rw` to the whitelist; list definitions merge with the
module's entry. Verified under the unit's sandbox (PrivateUsers,
PrivateTmp) that OpenCL init then succeeds as the jellyfin user.
2026-09-29 19:52:27 +08:00
Ryan Yin 2401727315 flake.lock: Update
Flake lock file updates:

• Updated input 'mysecrets':
    'git+ssh://git@github.com/ryan4yin/nix-secrets.git?ref=refs/heads/main&rev=b41ddcfa4537cfb56d1e0f0df0198a2e90fee180&shallow=1' (2026-09-25)
  → 'git+ssh://git@github.com/ryan4yin/nix-secrets.git?ref=refs/heads/main&rev=6daee5c0fa300bfcd13d702ae1d69aeb0a4c281d&shallow=1' (2026-09-28)
2026-09-28 16:19:41 +08:00
Ryan Yin 098ed110d7 chore(darwin): drop redundant broken-packages overlay
conda/mitmproxy/wireshark are already Linux-gated in the shared home modules, and verible is unreferenced, so replacing them with emptyDirectory is unnecessary.
2026-09-28 10:55:22 +08:00
Ryan Yin e8f8099011 chore(overlays): drop redundant openldap test override
openldap's flaky tests are skipped upstream now (doCheck = true), so the override only forced local rebuilds of openldap/cyrus-sasl dependents such as libvirt and kio.
2026-09-28 10:55:20 +08:00
Ryan Yin 470006ed3a chore(rustfs): drop nixpkgs-small input, use nixpkgs rustfs 1.0.0 2026-09-28 10:42:51 +08:00
Ryan Yin 5b4127bcec flake.lock: Update
Flake lock file updates:

• Updated input 'nixpkgs':
    'github:nixos/nixpkgs/e554fab72f81915600f3f449b786fd9af40439a5?narHash=sha256-ZKhUe/2IJUq1JhKxKMu8rbkgSGmPP2ZCqlIPn40aGCM%3D' (2026-09-17)
  → 'github:nixos/nixpkgs/e158d9ed9b51c98974c5e66e1ba1c9e0255fecaa?narHash=sha256-hKlVl12B1dF0Q5vd9dY3lIJM5mFGWYSlXwSLAqHZ1%2Bs%3D' (2026-09-26)
2026-09-28 10:41:47 +08:00
Ryan Yin b38d8b802e docs(agents): make apply a distinct step in impactful changes 2026-09-28 10:18:52 +08:00
Ryan Yin 853747425f chore(idols-ai): drop the davfs WebDAV fileshare mount
davfs2 caches the whole opened file regardless of cache_size, so opening media on /mnt/fileshare downloads it fully and blocks the mount. The share is no longer used.
2026-09-26 23:47:29 +08:00
ryan4yin | 二花 c1b590b31b chore(codex): use upstream 0.157.1 and drop the temporary patch (#327) 2026-09-26 23:24:39 +08:00
Ryan Yin 695d58de8b docs(agents): mention quoted heredoc for multi-line code 2026-09-26 22:53:13 +08:00
ryan4yin | 二花 1256570a84 feat(youko): firewall and tune transmission, stop homepage RPC probes (#326)
* fix(youko): stop homepage from probing the authenticated transmission RPC

The siteMonitor hit /transmission/rpc unauthenticated, so every probe returned 401 and counted against transmission's anti-brute-force threshold.

* feat(youko): tune transmission and firewall its network namespace

Raise the upload limit and queues, relax required encryption to preferred, and add a default-drop nftables input chain inside the netns so the peer port stays open while everything else is closed. Allowlist the nftables 'iif' keyword for typos.

* feat(youko): export transmission metrics and add a Grafana dashboard

nixpkgs has no transmission exporter, so add a small stdlib Python exporter that queries the authenticated netns RPC on loopback, scrape it from VictoriaMetrics, and ship a Transmission dashboard.

* fix(youko): cap transmission upload at 2500 kB/s

30 MB/s ignored the 50 Mbps uplink; keep seeding around 20 Mbps.

* Revert "feat(youko): export transmission metrics and add a Grafana dashboard"

This reverts commit 9df7a26526.
2026-09-26 22:40:02 +08:00
ryan4yin | 二花 e5b95affa6 feat(nixos): default-deny host inbound, trust the LAN/tailnet (#325)
Hosts used to trust the router as the only boundary (servers ran with the
firewall off; service ports were opened globally), so any LAN device could
reach every host service. Move filtering to the hosts as defence in depth
behind the router, which keeps its firewall on and uses a port mapping for
the few ports we actually need.

- base: nftables backend; deny inbound by default; trust the home LAN by
  source and the tailnet by interface (restrict the tailnet with ACLs);
  suppress per-service `openFirewall` global ports; allow node_exporter only
  from the monitoring host. The local podman bridge, the tailnet and the pod
  CIDRs are accepted only after the node_exporter drop.
- servers: stop disabling the firewall; drop redundant global port opens
  (caddy 80/443, NFS 2049, Jellyfin, localsend/test ports); delete the
  redundant desktop firewall module; aarch64/riscv64 import the firewall
  explicitly (akane was missing it).
- k3s: Cilium pods (10.0.0.0/8, fd00::/8) may reach host services; drop the
  flannel-era cni0/flannel.1 trust.
- youko: postgres listens on localhost + the LAN with pg_hba limited to the
  LAN; the NFS export stays LAN-scoped; remove the unauthenticated v2ray
  proxy.
- tests: `security-firewall` asserts every host runs the shared firewall, has
  the LAN CIDR rule and no global ports.
2026-09-26 14:02:03 +08:00
ryan4yin | 二花 2314754655 fix(youko): run transmission in its own network namespace (#324) 2026-09-26 12:48:51 +08:00
ryan4yin | 二花 f326dcb139 docs(agents): Simplifies global rules so agents can understand and follow it easily (#323) 2026-09-26 09:51:42 +08:00
Ryan Yin 27b99fa45f docs(agents): add untrusted-source safety rules
Treat content from unverified sources as data, not instructions, and require authorization before cloning, downloading, or executing unverified code and install scripts.
2026-09-26 08:23:15 +08:00
Ryan Yin 2a2933a889 docs(youko): note the transmission inbound-port TODO
youko's default gateway is suzi (mihomo transparent proxy, no UPnP/NAT-PMP),
so transmission cannot map its peer port; record the options for later.
2026-09-26 00:23:54 +08:00
ryan4yin | 二花 9fe6166fd1 feat(homepage): add Jellyfin, alerting, files and k3s-test-1 services (#322) 2026-09-26 00:15:06 +08:00
ryan4yin | 二花 a111676ede feat(youko): add Jellyfin media server (#321)
* feat(youko): add Jellyfin media server

Jellyfin reads the existing Transmission download tree on the public HDD
(via the shared `fileshare` group) and transcodes with the AMD Barceló
iGPU through VA-API. Encoding settings are declared in Nix
(`forceEncodingConfig`), so the web UI cannot silently drift.

Replace the per-service public ACME certificates with a single
`*.writefor.fun` wildcard (plus the apex) shared by every public vhost,
so adding a service needs no new ACME entry; a wildcard also keeps
individual hostnames out of public CT logs.

Pin the jellyfin uid/gid to avoid the dynamic-id drift fixed for the
other services in #320.

* fix(youko): enable graphics for Jellyfin VA-API

VA-API needs Mesa's radeonsi VA driver under /run/opengl-driver, but
youko is headless so hardware.graphics was off. Without it Jellyfin's
ffmpeg fails with 'va_openDriver() returns -1'. Enable it so the iGPU
can be used for hardware transcoding.

* fix(youko): use VA-API hardware encoding for Jellyfin

Enabling hardware decoding alone left encoding on the CPU (libx264), which
pins ~3 cores per 720p stream. Set transcoding.enableHardwareEncoding so
the iGPU (Vega VCN) does the H.264 encode too.

* feat(youko): harden Jellyfin proxy trust, groups, and HEVC encode

Add jellyfin to the render/video groups for robust VA-API access (was
relying on the world-writable render node), enable HEVC hardware encoding
(Vega's VCN supports it; AV1 does not), and patch network.xml on start to
trust Caddy as KnownProxies and advertise the public URL.

* fix(youko): run Jellyfin network.xml patch every start; allow LAN discovery

The module's preStart exits early once encoding.xml matches, so the
mkAfter network.xml patch only ran when encoding settings changed. Run it
with mkBefore instead. Also allow Jellyfin app auto-discovery (UDP 7359 /
TCP 8096) from the home network only.
2026-09-25 23:22:04 +08:00
ryan4yin | 二花 9187e4d920 fix(youko): pin dynamically-allocated service uid/gid (#320)
Service users/groups get a dynamically allocated uid/gid, which shifts when the user set changes: adding `immich` moved gitea 993:985 -> 996:993 and rustfs 987:977 -> 988:983. Files on the persistent volumes keep the old numeric owner, so gitea and rustfs lost access to their data and failed to start after a restart. Pin every dynamically allocated service id on youko to its current value. The ids are host-specific, so this is a host file.
2026-09-25 20:55:02 +08:00
ryan4yin | 二花 5ee8b728b5 fix(restic): use the correct systemd timer option names (#319)
timerConfig was set with lowercase onCalendar/randomizedDelaySec, which systemd treats as unknown directives. The generated timer had no schedule, so `systemctl list-timers` showed NEXT=- and no backup ever ran. Use the real keys OnCalendar/RandomizedDelaySec.
2026-09-25 20:39:59 +08:00
ryan4yin | 二花 4bd463a784 chore(eval): resolve catppuccin and rust-overlay deprecation warnings (#318) 2026-09-25 20:36:26 +08:00
ryan4yin | 二花 793b636fa5 feat(youko): add Immich photo platform (#317)
Add Immich via the native nixpkgs services.immich module on youko, reusing the host PostgreSQL 16 (pgvector + vectorchord) and a shared Valkey instance for Redis-compatible caching and BullMQ queues (Dragonfly rejected BullMQ's dynamic Lua keys). Media lives on the encrypted HDD at /data/apps/immich; Caddy publishes immich.writefor.fun with a publicly-trusted Let's Encrypt cert (security.acme + Cloudflare DNS-01) so the mobile app connects. OTEL metrics are scraped by VictoriaMetrics, and /var/cache is now preserved on the tmpfs-root hosts so the ML model cache survives reboots.
2026-09-25 20:30:01 +08:00
ryan4yin | 二花 6a6950e7eb fix(youko): don't redirect rustfs console STS/S3 calls (#316)
The s3-console Caddy vhost redirected every request to / to /rustfs/console/. The RustFS console frontend signs its own S3/STS requests to / (POST / for STS, GET /?x-id=ListBuckets), so the 302 broke the console session and login failed server-side with SignatureDoesNotMatch ("missing signed header: content-type"). Restrict the redirect to unsigned browser GET /; signed requests now reach RustFS unchanged.
2026-09-25 16:21:12 +08:00
ryan4yin | 二花 7d8d6ec8ad chore(youko): drop redundant loki-gateway Caddy vhost (#315) 2026-09-25 16:09:44 +08:00
ryan4yin | 二花 e8e7844051 Merge pull request #314 from ryan4yin/feat/rustfs-s3
feat(youko): replace MinIO with RustFS S3
2026-09-25 15:53:40 +08:00