Introduce /search command to assistant (#12372)

This pull request introduces semantic search to the assistant using a
slash command:


https://github.com/zed-industries/zed/assets/482957/62f39eae-d7d5-46bf-a356-dd081ff88312

Moreover, this also adds a status to pending slash commands, so that we
can show when a query is running or whether it failed:

<img width="1588" alt="image"
src="https://github.com/zed-industries/zed/assets/482957/e8d85960-6275-4552-a068-85efb74cfde1">

I think this could be better design-wise, but seems like a pretty good
start.

Release Notes:

- N/A

crates/semantic_index/src/embedding/cloud.rs

@@ -24,6 +24,10 @@ impl EmbeddingProvider for CloudEmbeddingProvider {
         // First, fetch any embeddings that are cached based on the requested texts' digests
         // Then compute any embeddings that are missing.
         async move {
+            if !self.client.status().borrow().is_connected() {
+                return Err(anyhow!("sign in required"));
+            }
+
             let cached_embeddings = self.client.request(proto::GetCachedEmbeddings {
                 model: self.model.clone(),
                 digests: texts