fix(nix): package local speech worker (#1587)

* fix(nix): package local speech worker

* fix(nix): reuse sherpa package helper

* fix(nix): patch sherpa-onnx prebuilt binaries for NixOS libstdc++

The prebuilt sherpa-onnx-linux-x64 .so files link against libstdc++.so.6
which is not on the NixOS library path. Add autoPatchelfHook to fix ELF
RPATHs and stdenv.cc.cc.lib to provide the C++ runtime, resolving the
"Failed to load model because protobuf parsing failed" SIGABRT at
speech worker startup.

Author: yhori991

nix/package.nix

@@ -5,6 +5,7 @@
   nodejs_22,
   python3,
   makeWrapper,
+  autoPatchelfHook,
   # node-pty needs libuv headers on Linux
   libuv,
   # Exposed so downstream flakes that follow a different nixpkgs revision
@@ -59,10 +60,13 @@ buildNpmPackage rec {
   nativeBuildInputs = [
     python3 # for node-gyp (node-pty compilation)
     makeWrapper
+  ] ++ lib.optionals stdenv.hostPlatform.isLinux [
+    autoPatchelfHook
   ];
 
   buildInputs = lib.optionals stdenv.hostPlatform.isLinux [
     libuv
+    stdenv.cc.cc.lib # libstdc++ for sherpa-onnx prebuilt binaries
   ];
 
   # Don't use the default npm build hook — we need a custom build sequence
@@ -71,10 +75,9 @@ buildNpmPackage rec {
   buildPhase = ''
     runHook preBuild
 
-    # Rebuild only node-pty (native addon for terminal emulation).
-    # Speech-related native modules (sherpa-onnx, onnxruntime-node) are
-    # intentionally left unbuilt — they're lazily loaded and gracefully
-    # degrade when unavailable.
+    # Rebuild only node-pty (native addon for terminal emulation). The sherpa
+    # speech runtime ships prebuilt platform packages and is copied into the
+    # daemon closure by scripts/trace-daemon.mjs.
     npm rebuild node-pty
 
     # Build all server packages in dependency order (defined in package.json)
@@ -89,7 +92,7 @@ buildNpmPackage rec {
 
     # Compute the daemon's runtime closure by static module-graph tracing
     # (@vercel/nft from supervisor-entrypoint.js, cli/dist/index.js, and the
-    # forked terminal-worker-process.js) plus an explicit list of non-JS
+    # forked terminal/speech worker processes) plus an explicit list of non-JS
     # assets read at runtime. The trace script is the single source of
     # truth for what the daemon needs at $out — auditable in plain JS, no
     # npm hoisting / .bin / workspace-symlink footguns.

scripts/trace-daemon.mjs

@@ -1,6 +1,6 @@
 #!/usr/bin/env node
 // Emit the set of files the daemon and CLI need at runtime, computed by
-// static module-graph tracing (@vercel/nft) from the three entry points.
+// static module-graph tracing (@vercel/nft) from the daemon entry points.
 // Used by nix/package.nix's installPhase to materialize $out/lib/paseo
 // with only the bytes the daemon actually loads — no Expo, RN, Metro,
 // Electron, ML stacks, or other non-daemon workspace bloat.
@@ -15,16 +15,27 @@
 import { nodeFileTrace } from "@vercel/nft";
 import { glob } from "node:fs/promises";
 import path from "node:path";
+import { pathToFileURL } from "node:url";
 
 const REPO_ROOT = path.resolve(import.meta.dirname, "..");
 
-// Three entry points. The terminal worker is forked into its own Node
-// process and has its own require tree (node-pty, etc.) — nft does not
-// follow fork boundaries, so it must be traced separately.
+const { sherpaPlatformPackageName } = await import(
+  pathToFileURL(
+    path.join(
+      REPO_ROOT,
+      "packages/server/dist/server/server/speech/providers/local/sherpa/sherpa-runtime-env.js",
+    ),
+  ).href
+);
+
+// Daemon entry points. Workers forked into their own Node processes have
+// independent require trees; nft does not follow fork boundaries, so trace
+// them separately.
 const entries = [
   "packages/cli/dist/index.js",
   "packages/server/dist/scripts/supervisor-entrypoint.js",
   "packages/server/dist/server/terminal/terminal-worker-process.js",
+  "packages/server/dist/server/server/speech/providers/local/worker-process.js",
 ];
 
 // Files read at runtime via fs APIs rather than `require`. nft only
@@ -44,19 +55,22 @@ const additionalInputs = [
   // the Nix derivation builds for one platform at a time and ships only
   // its own binaries.
   `node_modules/node-pty/prebuilds/${process.platform}-${process.arch}/**`,
+  // sherpa-onnx-node dynamically resolves a platform-specific native package.
+  // Copy the wrapper plus the host platform package explicitly.
+  "node_modules/sherpa-onnx-node/**",
+  `node_modules/${sherpaPlatformPackageName()}/**`,
 ];
 
 // Trace.
 const { fileList, warnings } = await nodeFileTrace(entries, {
   base: REPO_ROOT,
   // Tolerate the conditional / dynamic patterns we already audited:
-  // sherpa-onnx-${platform}-${arch} package resolution (sherpa is
-  // intentionally not built in the Nix sandbox; voice features degrade
-  // gracefully when unavailable), and a handful of test-only requires
-  // that get tree-shaken out by tsc.
+  // sherpa-onnx-${platform}-${arch} package resolution (the host package
+  // is copied explicitly above), and a handful of test-only requires that
+  // get tree-shaken out by tsc.
   ignore: [
     // Cross-platform native packages for the sherpa speech runtime;
-    // unsupported in the Nix build, lazily loaded when present.
+    // only the host platform package is needed at runtime.
     "sherpa-onnx-*/**",
     // Platform-specific clipboard variants; only the host's variant
     // is needed at runtime, and the package's index.js resolver picks