Initialize repository with harness code and assets

Initial import of all source code, config, and README assets: the
packages workspace (cli, core, server, web, docs, landing, skills),
build scripts, tooling config, and CI workflows.

Includes the data-layout revision made on this branch: the local data
root defaults to ~/.penguin/data (PENGUIN_HOME still overrides; the
installer keeps its binaries in ~/.penguin), and every Agent lives
under <project>/agents/<agent>/ — path helpers, the three
agent-enumeration scans, the system prompt, built-in Skills, tests
and docs all follow the new layout.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018ihk8iQuo3kv2aPjAYEPuR

packages/web/src/lib/api-error.ts

@@ -0,0 +1,25 @@
+/**
+ * API error → display text: the server's message is hardcoded in Chinese
+ * (`{error:{code,message}}`), but the UI language may be English — so for any
+ * error the user might actually hit, we derive localized text from **code**;
+ * the server message is only a fallback for unknown codes.
+ */
+import { ApiError } from "../api/client";
+import { S } from "./strings";
+
+/**
+ * ctx supplies context needed for localized text (e.g. the model id for a
+ * missing key — the error body only has the code; the model id is filled in
+ * by the caller: the model used by the current session/draft when sending).
+ */
+export function apiErrorText(err: unknown, ctx?: { modelId?: string }): string {
+  if (!(err instanceof ApiError)) return S.common.unknownError;
+  switch (err.code) {
+    case "model_credential_missing":
+      return ctx?.modelId ? S.errors.modelCredentialMissing(ctx.modelId) : err.message;
+    case "no_default_model":
+      return S.errors.noDefaultModel;
+    default:
+      return err.message;
+  }
+}