Add the harness, envs, per-bot models and chat attachments

- Harness: providers, model catalog synced hourly, automatic free picks,
  cooldowns, tool loop with context trimming, MCP computer tools
- Bots queue runs replies and tasks through the harness; jobs wait
  (attempt refunded) while every model is resting
- Envs: Settings > Environment, scoped general/workspace/computer,
  written to computers and redacted from tool output
- Models: Settings > Models, per-harness provider select defaulting to
  OpenRouter, and a models modal on each bot's chat (the coordinator's
  edits the defaults)
- Chat: thinking row while a reply runs, file attachments (upload,
  paste, drop) that replies read as images, audio and document text
- Fix media uploads failing after a user's first file, and private
  files being served from the public media path

Co-Authored-By: Claude Opus 5.5 <[email protected]>

src/harness/providers/openai-compatible/transcribe.ts

@@ -0,0 +1,42 @@
+import type { HarnessAdapter } from "../../types";
+import providerRequest from "./provider-request";
+
+/**
+ * `POST /audio/transcriptions` as an OpenAI-style file upload, which
+ * OpenRouter, OpenAI, Groq and local servers all take.
+ */
+const transcribe: HarnessAdapter["transcribe"] = async ({
+    candidate,
+    audio,
+    language,
+    signal,
+}) => {
+    const form = new FormData();
+
+    form.append("model", candidate.model.slug);
+    form.append(
+        "file",
+        new Blob([new Uint8Array(audio.data)], {
+            type: `audio/${audio.format}`,
+        }),
+        `audio.${audio.format}`,
+    );
+
+    if (language) {
+        form.append("language", language);
+    }
+
+    const res = await providerRequest({
+        provider: candidate.provider,
+        path: "/audio/transcriptions",
+        form,
+        signal,
+        candidate_key: candidate.key,
+    });
+
+    const json = await res.json().catch(() => null);
+
+    return { text: String(json?.text || "").trim() };
+};
+
+export default transcribe;