chore: refactor retrieval pipeline to chunk-first RRF with derived entities and slimmer eval surface.

Collapse the multi-strategy entity engine into one benchmarked chunk retrieval path, derive entities from retrieved chunks, and update consumers, docs, and clippy fixes across the workspace.

retrieval-pipeline/Cargo.toml

@@ -10,16 +10,14 @@ workspace = true
 [dependencies]
 tokio = { workspace = true }
 serde = { workspace = true }
-axum = { workspace = true }
 tracing = { workspace = true }
-anyhow = { workspace = true }
-thiserror = { workspace = true }
 serde_json = { workspace = true }
-surrealdb = { workspace = true }
-futures = { workspace = true }
 async-openai = { workspace = true }
 async-trait = { workspace = true }
-uuid = { workspace = true }
 fastembed = { workspace = true }
 
 common = { path = "../common", features = ["test-utils"] }
+
+[dev-dependencies]
+anyhow = { workspace = true }
+uuid = { workspace = true }