feat(packaging): core distribution packages (deb + rpm + tar.gz) via nfpm

Previously only the Electron GUI had installers; server admins had to build the
core by hand (`make build` -> bare binary). This adds a first-class server
distribution:

- `make core-dist` -> packaging/core-dist.sh -> cmd/build/dist/
    llmsproxy_<ver>_amd64.deb        (systemd unit + adapters + example config)
    llmsproxy-<ver>.x86_64.rpm
    llmsproxy-<ver>-linux-amd64.tar.gz
- nfpm config (packaging/nfpm.yaml): binary to /usr/bin, systemd unit with the
  measured memory tuning (GOGC=50, MALLOC_ARENA_MAX=2) baked in, Lua adapters to
  /usr/share/llmsproxy/adapters, repo's config.yaml as the example.
- postinst seeds /etc/llmsproxy/config.yaml on first install and keeps existing
  runtime state across reinstalls; prerm stops the service on removal.

Gotchas handled: nfpm v2 here does not render `{{ .Env.VERSION }}`, so the script
stamps the version into a throwaway config copy; and under `set -o pipefail` the
idiom `strings | grep -q` is broken (grep -q closes the pipe and strings dies on
SIGPIPE -> spurious 141), so the symbol sanity-gate stages strings in a temp file.

README (zh/en) updated: memory figures replaced with production-measured
32-35 MB settled (was 37-42), and the packaging docs now describe core-dist and
the dockerized Windows build.

packaging/nfpm.yaml

@@ -0,0 +1,59 @@
+# nfpm config for the llmsproxy core gateway binary.
+# The version placeholder below is substituted by packaging/core-dist.sh (nfpm v2
+# here does not honour {{ .Env.VERSION }} template rendering), so do not edit it
+# by hand.
+name: llmsproxy
+arch: amd64
+platform: linux
+version: __VERSION__
+epoch: "1"
+section: net
+priority: optional
+maintainer: JianFeeeee <dev@modelrouter.local>
+description: |
+  Unified OpenAI-compatible multi-source LLM gateway.
+  Routes requests across multiple upstream LLM providers with Lua adapters,
+  model scheduling, quota windows, an elastic Lua worker pool, and a Web UI.
+vendor: ModelRouter
+homepage: https://gitcode.com/JianFeeeee/ModelRouter
+license: MIT
+contents:
+  # binary
+  - src: ./cmd/build/llmsproxy
+    dst: /usr/bin/llmsproxy
+    file_info:
+      mode: 0755
+  # systemd unit (memory tuning baked in)
+  - src: ./packaging/llmsproxy.service
+    dst: /etc/systemd/system/llmsproxy.service
+    file_info:
+      mode: 0644
+  # example config (seeded to /etc/llmsproxy on first install by postinst)
+  - src: ./config.yaml
+    dst: /usr/share/llmsproxy/config.example.yaml
+    file_info:
+      mode: 0644
+  # built-in Lua adapters (deployed to /etc/llmsproxy/adapters)
+  - src: ./internal/lua/adapters/*.lua
+    dst: /usr/share/llmsproxy/adapters/
+    file_info:
+      mode: 0644
+  # README pointer
+  - src: ./README.md
+    dst: /usr/share/doc/llmsproxy/README.md
+    file_info:
+      mode: 0644
+
+scripts:
+  postinstall: ./packaging/postinst.sh
+  preremove: ./packaging/prerm.sh
+
+rpm:
+  group: Applications/Internet
+  summary: Unified OpenAI-compatible multi-source LLM gateway
+  compression: gzip
+
+deb:
+  compression: gzip
+  fields:
+    Recommends: ca-certificates