refactor(rag): drop the hardcoded embedding model hint from attach

The attach wizard mapped a collection's vector dimension to a hardcoded list of
model names and printed them as likely candidates. The list was never checked
against the models the user actually has configured, so it could recommend a
model they cannot select, and one entry was a parenthetical note rather than a
model id and so could never match anything. Any list like this rots as models
are released.

The dimension itself comes from the server and is worth stating, so it is still
printed, as is the warning that a mismatched embedding model returns bad
results. Deriving real candidates would need a dimension recorded against each
configured model, which the model config does not carry today.

src/rag/mod.rs

@@ -491,13 +491,7 @@ impl Rag {
             );
         }
         if dim > 0 {
-            let candidates = embedding_model_candidates_for_dimension(dim);
+            println!("Collection uses {dim}-dim vectors.");
-            if !candidates.is_empty() {
-                println!(
-                    "Collection uses {dim}-dim vectors. Likely models: {}",
-                    candidates.join(", ")
-                );
-            }
         }
         println!(
             "⚠️  If the embedding model doesn't match what built this collection, \
@@ -1834,21 +1828,6 @@ fn driver_auth_header(driver: &str) -> (&'static str, &'static str) {
     }
 }
 
-/// Embedding models known to produce a given vector dimension, used to hint the
-/// user toward a model compatible with the collection they just picked.
-fn embedding_model_candidates_for_dimension(dim: u64) -> Vec<&'static str> {
-    match dim {
-        1536 => vec!["text-embedding-3-small", "text-embedding-ada-002"],
-        3072 => vec!["text-embedding-3-large"],
-        768 => vec!["nomic-embed-text", "all-minilm-l6-v2"],
-        1024 => vec![
-            "text-embedding-3-small (matryoshka-1024)",
-            "jina-embeddings-v2-base",
-        ],
-        _ => vec![],
-    }
-}
-
 fn select_embedding_model(models: &[&Model]) -> Result<String> {
     let max_width = models.iter().map(|v| v.id().len()).max().unwrap_or(0);
     let models: Vec<_> = models