From 4b0535f85202bed3885a4988232f194c8c7d9a6c Mon Sep 17 00:00:00 2001
From: teernisse <teernisse@visiostack.com>
Date: Fri, 6 Mar 2026 17:07:17 -0500
Subject: [PATCH] perf(timeline): guard against overly broad seed queries

Add pre-flight FTS count check before expensive bm25-ranked search.
Queries matching >10,000 documents are rejected instantly with a
suggestion to use a more specific query or --since filter.

Prevents multi-minute CPU spin on queries like 'merge request' that
match most of the corpus (106K/178K documents).
---
 src/timeline/seed.rs | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/src/timeline/seed.rs b/src/timeline/seed.rs
index caa44d0..fd26e31 100644
--- a/src/timeline/seed.rs
+++ b/src/timeline/seed.rs
@@ -48,6 +48,21 @@ pub async fn seed_timeline(
         });
     }
 
+    // Guard: reject overly broad queries before running expensive ranked search.
+    // The count query (no bm25/snippet) is cheap even on broad matches.
+    const SEED_MATCH_CEILING: i64 = 10_000;
+    let match_count: i64 = conn.query_row(
+        "SELECT COUNT(*) FROM documents_fts WHERE documents_fts MATCH ?1",
+        [&fts_query],
+        |row| row.get(0),
+    )?;
+    if match_count > SEED_MATCH_CEILING {
+        return Err(crate::core::error::LoreError::Other(format!(
+            "Query too broad: matched {match_count} documents (ceiling: {SEED_MATCH_CEILING}). \
+             Use a more specific query or narrow with --since.",
+        )));
+    }
+
     // Use hybrid search for seed entity discovery (better recall than FTS alone).
     // search_hybrid gracefully falls back to FTS-only when Ollama is unavailable.
     let filters = SearchFilters {