chore: add gitignore for mock-seed, roam CI workflow, formatting

- Add tools/mock-seed/ to .gitignore
- Add .github/workflows/roam.yml CI workflow
- Add .roam/fitness.yaml architectural fitness rules
- Rustfmt formatting fixes in show.rs and vector.rs
- Beads sync

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

.github/workflows/roam.yml

@@ -0,0 +1,21 @@
+name: Roam Code Analysis
+on:
+  pull_request:
+    branches: [main, master]
+permissions:
+  contents: read
+  pull-requests: write
+jobs:
+  roam:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+      - run: pip install roam-code
+      - run: roam index
+      - run: roam fitness
+      - run: roam pr-risk --json

.gitignore

@@ -41,6 +41,9 @@ lore.config.json
 *.db-shm
 
 
+# Mock seed data
+tools/mock-seed/
+
 # Added by cargo
 
 /target

.roam/fitness.yaml

@@ -0,0 +1,11 @@
+rules:
+  - name: No circular imports in core
+    type: dependency
+    source: "src/**"
+    forbidden_target: "tests/**"
+    reason: "Production code should not import test modules"
+  - name: Complexity threshold
+    type: metric
+    metric: cognitive_complexity
+    threshold: 30
+    reason: "Functions above 30 cognitive complexity need refactoring"

Cargo.lock

@@ -1106,7 +1106,7 @@ checksum = "5e5032e24019045c762d3c0f28f5b6b8bbf38563a65908389bf7978758920897"
 
 [[package]]
 name = "lore"
-version = "0.8.0"
+version = "0.8.1"
 dependencies = [
  "async-stream",
  "chrono",

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "lore"
-version = "0.8.0"
+version = "0.8.1"
 edition = "2024"
 description = "Gitlore - Local GitLab data management with semantic search"
 authors = ["Taylor Eernisse"]

src/cli/commands/list.rs

@@ -6,6 +6,7 @@ use crate::Config;
 use crate::cli::robot::{RobotMeta, expand_fields_preset, filter_fields};
 use crate::core::db::create_connection;
 use crate::core::error::{LoreError, Result};
+use crate::core::path_resolver::escape_like as note_escape_like;
 use crate::core::paths::get_db_path;
 use crate::core::project::resolve_project;
 use crate::core::time::{ms_to_iso, now_ms, parse_since};
@@ -1257,13 +1258,6 @@ pub struct NoteListFilters {
     pub order: String,
 }
 
-fn note_escape_like(input: &str) -> String {
-    input
-        .replace('\\', "\\\\")
-        .replace('%', "\\%")
-        .replace('_', "\\_")
-}
-
 pub fn query_notes(
     conn: &Connection,
     filters: &NoteListFilters,

src/cli/commands/show.rs

@@ -160,6 +160,7 @@ pub fn run_show_issue(
     })
 }
 
+#[derive(Debug)]
 struct IssueRow {
     id: i64,
     iid: i64,
@@ -194,7 +195,7 @@ fn find_issue(conn: &Connection, iid: i64, project_filter: Option<&str>) -> Resu
                         i.due_date, i.milestone_title,
                         (SELECT COUNT(*) FROM notes n
                          JOIN discussions d ON n.discussion_id = d.id
-                         WHERE d.noteable_type = 'Issue' AND d.noteable_id = i.id AND n.is_system = 0) AS user_notes_count,
+                         WHERE d.noteable_type = 'Issue' AND d.issue_id = i.id AND n.is_system = 0) AS user_notes_count,
                         i.status_name, i.status_category, i.status_color,
                         i.status_icon_name, i.status_synced_at
                  FROM issues i
@@ -210,7 +211,7 @@ fn find_issue(conn: &Connection, iid: i64, project_filter: Option<&str>) -> Resu
                     i.due_date, i.milestone_title,
                     (SELECT COUNT(*) FROM notes n
                      JOIN discussions d ON n.discussion_id = d.id
-                     WHERE d.noteable_type = 'Issue' AND d.noteable_id = i.id AND n.is_system = 0) AS user_notes_count,
+                     WHERE d.noteable_type = 'Issue' AND d.issue_id = i.id AND n.is_system = 0) AS user_notes_count,
                     i.status_name, i.status_category, i.status_color,
                     i.status_icon_name, i.status_synced_at
              FROM issues i
@@ -1218,6 +1219,172 @@ mod tests {
         .unwrap();
     }
 
+    fn seed_second_project(conn: &Connection) {
+        conn.execute(
+            "INSERT INTO projects (id, gitlab_project_id, path_with_namespace, web_url, created_at, updated_at)
+             VALUES (2, 101, 'other/repo', 'https://gitlab.example.com/other', 1000, 2000)",
+            [],
+        )
+        .unwrap();
+    }
+
+    fn seed_discussion_with_notes(
+        conn: &Connection,
+        issue_id: i64,
+        project_id: i64,
+        user_notes: usize,
+        system_notes: usize,
+    ) {
+        let disc_id: i64 = conn
+            .query_row(
+                "SELECT COALESCE(MAX(id), 0) + 1 FROM discussions",
+                [],
+                |r| r.get(0),
+            )
+            .unwrap();
+        conn.execute(
+            "INSERT INTO discussions (id, gitlab_discussion_id, project_id, issue_id, noteable_type, first_note_at, last_note_at, last_seen_at)
+             VALUES (?1, ?2, ?3, ?4, 'Issue', 1000, 2000, 2000)",
+            rusqlite::params![disc_id, format!("disc-{}", disc_id), project_id, issue_id],
+        )
+        .unwrap();
+        for i in 0..user_notes {
+            conn.execute(
+                "INSERT INTO notes (gitlab_id, discussion_id, project_id, author_username, body, created_at, updated_at, last_seen_at, is_system, position)
+                 VALUES (?1, ?2, ?3, 'user1', 'comment', 1000, 2000, 2000, 0, ?4)",
+                rusqlite::params![1000 + disc_id * 100 + i as i64, disc_id, project_id, i as i64],
+            )
+            .unwrap();
+        }
+        for i in 0..system_notes {
+            conn.execute(
+                "INSERT INTO notes (gitlab_id, discussion_id, project_id, author_username, body, created_at, updated_at, last_seen_at, is_system, position)
+                 VALUES (?1, ?2, ?3, 'system', 'status changed', 1000, 2000, 2000, 1, ?4)",
+                rusqlite::params![2000 + disc_id * 100 + i as i64, disc_id, project_id, (user_notes + i) as i64],
+            )
+            .unwrap();
+        }
+    }
+
+    // --- find_issue tests ---
+
+    #[test]
+    fn test_find_issue_basic() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        let row = find_issue(&conn, 10, None).unwrap();
+        assert_eq!(row.iid, 10);
+        assert_eq!(row.title, "Test issue");
+        assert_eq!(row.state, "opened");
+        assert_eq!(row.author_username, "author");
+        assert_eq!(row.project_path, "group/repo");
+    }
+
+    #[test]
+    fn test_find_issue_with_project_filter() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        let row = find_issue(&conn, 10, Some("group/repo")).unwrap();
+        assert_eq!(row.iid, 10);
+        assert_eq!(row.project_path, "group/repo");
+    }
+
+    #[test]
+    fn test_find_issue_not_found() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        let err = find_issue(&conn, 999, None).unwrap_err();
+        assert!(matches!(err, LoreError::NotFound(_)));
+    }
+
+    #[test]
+    fn test_find_issue_wrong_project_filter() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        seed_second_project(&conn);
+        // Issue 10 only exists in project 1, not project 2
+        let err = find_issue(&conn, 10, Some("other/repo")).unwrap_err();
+        assert!(matches!(err, LoreError::NotFound(_)));
+    }
+
+    #[test]
+    fn test_find_issue_ambiguous_without_project() {
+        let conn = setup_test_db();
+        seed_issue(&conn); // issue iid=10 in project 1
+        seed_second_project(&conn);
+        conn.execute(
+            "INSERT INTO issues (id, gitlab_id, iid, project_id, title, state, author_username,
+             created_at, updated_at, last_seen_at)
+             VALUES (2, 201, 10, 2, 'Same iid different project', 'opened', 'author', 1000, 2000, 2000)",
+            [],
+        )
+        .unwrap();
+        let err = find_issue(&conn, 10, None).unwrap_err();
+        assert!(matches!(err, LoreError::Ambiguous(_)));
+    }
+
+    #[test]
+    fn test_find_issue_ambiguous_resolved_with_project() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        seed_second_project(&conn);
+        conn.execute(
+            "INSERT INTO issues (id, gitlab_id, iid, project_id, title, state, author_username,
+             created_at, updated_at, last_seen_at)
+             VALUES (2, 201, 10, 2, 'Same iid different project', 'opened', 'author', 1000, 2000, 2000)",
+            [],
+        )
+        .unwrap();
+        let row = find_issue(&conn, 10, Some("other/repo")).unwrap();
+        assert_eq!(row.title, "Same iid different project");
+    }
+
+    #[test]
+    fn test_find_issue_user_notes_count_zero() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        let row = find_issue(&conn, 10, None).unwrap();
+        assert_eq!(row.user_notes_count, 0);
+    }
+
+    #[test]
+    fn test_find_issue_user_notes_count_excludes_system() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        // 2 user notes + 3 system notes = should count only 2
+        seed_discussion_with_notes(&conn, 1, 1, 2, 3);
+        let row = find_issue(&conn, 10, None).unwrap();
+        assert_eq!(row.user_notes_count, 2);
+    }
+
+    #[test]
+    fn test_find_issue_user_notes_count_across_discussions() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        seed_discussion_with_notes(&conn, 1, 1, 3, 0); // 3 user notes
+        seed_discussion_with_notes(&conn, 1, 1, 1, 2); // 1 user note + 2 system
+        let row = find_issue(&conn, 10, None).unwrap();
+        assert_eq!(row.user_notes_count, 4);
+    }
+
+    #[test]
+    fn test_find_issue_notes_count_ignores_other_issues() {
+        let conn = setup_test_db();
+        seed_issue(&conn);
+        // Add a second issue
+        conn.execute(
+            "INSERT INTO issues (id, gitlab_id, iid, project_id, title, state, author_username,
+             created_at, updated_at, last_seen_at)
+             VALUES (2, 201, 20, 1, 'Other issue', 'opened', 'author', 1000, 2000, 2000)",
+            [],
+        )
+        .unwrap();
+        // Notes on issue 2, not issue 1
+        seed_discussion_with_notes(&conn, 2, 1, 5, 0);
+        let row = find_issue(&conn, 10, None).unwrap();
+        assert_eq!(row.user_notes_count, 0); // Issue 10 has no notes
+    }
+
     #[test]
     fn test_ansi256_from_rgb() {
         assert_eq!(ansi256_from_rgb(0, 0, 0), 16);

src/cli/commands/timeline.rs

@@ -13,6 +13,7 @@ use crate::core::timeline::{
 use crate::core::timeline_collect::collect_events;
 use crate::core::timeline_expand::expand_timeline;
 use crate::core::timeline_seed::seed_timeline;
+use crate::embedding::ollama::{OllamaClient, OllamaConfig};
 
 /// Parameters for running the timeline pipeline.
 pub struct TimelineParams {
@@ -28,7 +29,7 @@ pub struct TimelineParams {
 }
 
 /// Run the full timeline pipeline: SEED -> EXPAND -> COLLECT.
-pub fn run_timeline(config: &Config, params: &TimelineParams) -> Result<TimelineResult> {
+pub async fn run_timeline(config: &Config, params: &TimelineParams) -> Result<TimelineResult> {
     let db_path = get_db_path(config.storage.db_path.as_deref());
     let conn = create_connection(&db_path)?;
 
@@ -50,15 +51,25 @@ pub fn run_timeline(config: &Config, params: &TimelineParams) -> Result<Timeline
         })
         .transpose()?;
 
-    // Stage 1+2: SEED + HYDRATE
+    // Construct OllamaClient for hybrid search (same pattern as run_search)
+    let ollama_cfg = &config.embedding;
+    let client = OllamaClient::new(OllamaConfig {
+        base_url: ollama_cfg.base_url.clone(),
+        model: ollama_cfg.model.clone(),
+        ..OllamaConfig::default()
+    });
+
+    // Stage 1+2: SEED + HYDRATE (hybrid search with FTS fallback)
     let seed_result = seed_timeline(
         &conn,
+        Some(&client),
         &params.query,
         project_id,
         since_ms,
         params.max_seeds,
         params.max_evidence,
-    )?;
+    )
+    .await?;
 
     // Stage 3: EXPAND
     let expand_result = expand_timeline(
@@ -81,6 +92,7 @@ pub fn run_timeline(config: &Config, params: &TimelineParams) -> Result<Timeline
 
     Ok(TimelineResult {
         query: params.query.clone(),
+        search_mode: seed_result.search_mode,
         events,
         total_events_before_limit: total_before_limit,
         seed_entities: seed_result.seed_entities,
@@ -258,7 +270,7 @@ pub fn print_timeline_json_with_meta(
         ok: true,
         data: TimelineDataJson::from_result(result),
         meta: TimelineMetaJson {
-            search_mode: "lexical".to_owned(),
+            search_mode: result.search_mode.clone(),
             expansion_depth: depth,
             expand_mentions,
             total_entities: result.seed_entities.len() + result.expanded_entities.len(),

src/cli/commands/who.rs

@@ -9,6 +9,9 @@ use crate::cli::robot::RobotMeta;
 use crate::core::config::ScoringConfig;
 use crate::core::db::create_connection;
 use crate::core::error::{LoreError, Result};
+use crate::core::path_resolver::{PathQuery, build_path_query, normalize_repo_path};
+#[cfg(test)]
+use crate::core::path_resolver::{SuffixResult, escape_like, suffix_probe};
 use crate::core::paths::get_db_path;
 use crate::core::project::resolve_project;
 use crate::core::time::{ms_to_iso, now_ms, parse_since, parse_since_from};
@@ -75,30 +78,6 @@ fn resolve_mode<'a>(args: &'a WhoArgs) -> Result<WhoMode<'a>> {
     ))
 }
 
-/// Normalize user-supplied repo paths to match stored DiffNote paths.
-/// - trims whitespace
-/// - strips leading "./" and "/" (repo-relative paths)
-/// - converts '\' to '/' when no '/' present (Windows paste)
-/// - collapses repeated "//"
-fn normalize_repo_path(input: &str) -> String {
-    let mut s = input.trim().to_string();
-    // Windows backslash normalization (only when no forward slashes present)
-    if s.contains('\\') && !s.contains('/') {
-        s = s.replace('\\', "/");
-    }
-    // Strip leading ./
-    while s.starts_with("./") {
-        s = s[2..].to_string();
-    }
-    // Strip leading /
-    s = s.trim_start_matches('/').to_string();
-    // Collapse repeated //
-    while s.contains("//") {
-        s = s.replace("//", "/");
-    }
-    s
-}
-
 // ─── Result Types ────────────────────────────────────────────────────────────
 
 /// Top-level run result: carries resolved inputs + the mode-specific result.
@@ -485,210 +464,6 @@ fn resolve_since_required(input: &str) -> Result<i64> {
 
 // ─── Path Query Construction ─────────────────────────────────────────────────
 
-/// Describes how to match a user-supplied path in SQL.
-#[derive(Debug)]
-struct PathQuery {
-    /// The parameter value to bind.
-    value: String,
-    /// If true: use `LIKE value ESCAPE '\'`. If false: use `= value`.
-    is_prefix: bool,
-}
-
-/// Build a path query from a user-supplied path, with project-scoped DB probes.
-///
-/// Rules:
-/// - If the path ends with `/`, it's a directory prefix -> `escaped_path/%` (LIKE)
-/// - If the path is a root path (no `/`) and does NOT end with `/`, treat as exact (=)
-/// - Else if the last path segment contains `.`, heuristic suggests file (=)
-/// - Two-way DB probe (project-scoped): when heuristics are ambiguous,
-///   probe the DB to resolve.
-/// - Otherwise, treat as directory prefix -> `escaped_path/%` (LIKE)
-fn build_path_query(conn: &Connection, path: &str, project_id: Option<i64>) -> Result<PathQuery> {
-    let trimmed = path.trim_end_matches('/');
-    let last_segment = trimmed.rsplit('/').next().unwrap_or(trimmed);
-    let is_root = !trimmed.contains('/');
-    let forced_dir = path.ends_with('/');
-    // Heuristic is now only a fallback; probes decide first when ambiguous.
-    let looks_like_file = !forced_dir && (is_root || last_segment.contains('.'));
-
-    // Probe 1: exact file exists in DiffNotes OR mr_file_changes (project-scoped)
-    // Checks both new_path and old_path to support querying renamed files.
-    // Exact-match probes already use the partial index, but LIKE probes below
-    // benefit from the INDEXED BY hint (same planner issue as expert query).
-    let exact_exists = conn
-        .query_row(
-            "SELECT 1 FROM notes INDEXED BY idx_notes_diffnote_path_created
-         WHERE note_type = 'DiffNote'
-           AND is_system = 0
-           AND (position_new_path = ?1 OR position_old_path = ?1)
-           AND (?2 IS NULL OR project_id = ?2)
-         LIMIT 1",
-            rusqlite::params![trimmed, project_id],
-            |_| Ok(()),
-        )
-        .is_ok()
-        || conn
-            .query_row(
-                "SELECT 1 FROM mr_file_changes
-             WHERE (new_path = ?1 OR old_path = ?1)
-               AND (?2 IS NULL OR project_id = ?2)
-             LIMIT 1",
-                rusqlite::params![trimmed, project_id],
-                |_| Ok(()),
-            )
-            .is_ok();
-
-    // Probe 2: directory prefix exists in DiffNotes OR mr_file_changes (project-scoped)
-    // Checks both new_path and old_path to support querying renamed directories.
-    let prefix_exists = if !forced_dir && !exact_exists {
-        let escaped = escape_like(trimmed);
-        let pat = format!("{escaped}/%");
-        conn.query_row(
-            "SELECT 1 FROM notes INDEXED BY idx_notes_diffnote_path_created
-             WHERE note_type = 'DiffNote'
-               AND is_system = 0
-               AND (position_new_path LIKE ?1 ESCAPE '\\' OR position_old_path LIKE ?1 ESCAPE '\\')
-               AND (?2 IS NULL OR project_id = ?2)
-             LIMIT 1",
-            rusqlite::params![pat, project_id],
-            |_| Ok(()),
-        )
-        .is_ok()
-            || conn
-                .query_row(
-                    "SELECT 1 FROM mr_file_changes
-                 WHERE (new_path LIKE ?1 ESCAPE '\\' OR old_path LIKE ?1 ESCAPE '\\')
-                   AND (?2 IS NULL OR project_id = ?2)
-                 LIMIT 1",
-                    rusqlite::params![pat, project_id],
-                    |_| Ok(()),
-                )
-                .is_ok()
-    } else {
-        false
-    };
-
-    // Probe 3: suffix match — user typed a bare filename or partial path that
-    // doesn't exist as-is. Search for full paths ending with /input (or equal to input).
-    // This handles "login.rs" matching "src/auth/login.rs".
-    let suffix_resolved = if !forced_dir && !exact_exists && !prefix_exists && looks_like_file {
-        suffix_probe(conn, trimmed, project_id)?
-    } else {
-        SuffixResult::NotAttempted
-    };
-
-    match suffix_resolved {
-        SuffixResult::Unique(full_path) => Ok(PathQuery {
-            value: full_path,
-            is_prefix: false,
-        }),
-        SuffixResult::Ambiguous(candidates) => {
-            let list = candidates
-                .iter()
-                .map(|p| format!("  {p}"))
-                .collect::<Vec<_>>()
-                .join("\n");
-            Err(LoreError::Ambiguous(format!(
-                "'{trimmed}' matches multiple paths. Use the full path or -p to scope:\n{list}"
-            )))
-        }
-        SuffixResult::NotAttempted | SuffixResult::NoMatch => {
-            // Original logic: exact > prefix > heuristic
-            let is_file = if forced_dir {
-                false
-            } else if exact_exists {
-                true
-            } else if prefix_exists {
-                false
-            } else {
-                looks_like_file
-            };
-
-            if is_file {
-                Ok(PathQuery {
-                    value: trimmed.to_string(),
-                    is_prefix: false,
-                })
-            } else {
-                let escaped = escape_like(trimmed);
-                Ok(PathQuery {
-                    value: format!("{escaped}/%"),
-                    is_prefix: true,
-                })
-            }
-        }
-    }
-}
-
-/// Result of a suffix probe against the DB.
-enum SuffixResult {
-    /// Suffix probe was not attempted (conditions not met).
-    NotAttempted,
-    /// No paths matched the suffix.
-    NoMatch,
-    /// Exactly one distinct path matched — auto-resolve.
-    Unique(String),
-    /// Multiple distinct paths matched — user must disambiguate.
-    Ambiguous(Vec<String>),
-}
-
-/// Probe both notes and mr_file_changes for paths ending with the given suffix.
-/// Searches both new_path and old_path columns to support renamed file resolution.
-/// Returns up to 11 distinct candidates (enough to detect ambiguity + show a useful list).
-fn suffix_probe(conn: &Connection, suffix: &str, project_id: Option<i64>) -> Result<SuffixResult> {
-    let escaped = escape_like(suffix);
-    let suffix_pat = format!("%/{escaped}");
-
-    let mut stmt = conn.prepare_cached(
-        "SELECT DISTINCT full_path FROM (
-            SELECT position_new_path AS full_path
-            FROM notes INDEXED BY idx_notes_diffnote_path_created
-            WHERE note_type = 'DiffNote'
-              AND is_system = 0
-              AND (position_new_path LIKE ?1 ESCAPE '\\' OR position_new_path = ?2)
-              AND (?3 IS NULL OR project_id = ?3)
-            UNION
-            SELECT new_path AS full_path FROM mr_file_changes
-            WHERE (new_path LIKE ?1 ESCAPE '\\' OR new_path = ?2)
-              AND (?3 IS NULL OR project_id = ?3)
-            UNION
-            SELECT position_old_path AS full_path FROM notes
-            WHERE note_type = 'DiffNote'
-              AND is_system = 0
-              AND position_old_path IS NOT NULL
-              AND (position_old_path LIKE ?1 ESCAPE '\\' OR position_old_path = ?2)
-              AND (?3 IS NULL OR project_id = ?3)
-            UNION
-            SELECT old_path AS full_path FROM mr_file_changes
-            WHERE old_path IS NOT NULL
-              AND (old_path LIKE ?1 ESCAPE '\\' OR old_path = ?2)
-              AND (?3 IS NULL OR project_id = ?3)
-        )
-        ORDER BY full_path
-        LIMIT 11",
-    )?;
-
-    let candidates: Vec<String> = stmt
-        .query_map(rusqlite::params![suffix_pat, suffix, project_id], |row| {
-            row.get(0)
-        })?
-        .collect::<std::result::Result<Vec<_>, _>>()?;
-
-    match candidates.len() {
-        0 => Ok(SuffixResult::NoMatch),
-        1 => Ok(SuffixResult::Unique(candidates.into_iter().next().unwrap())),
-        _ => Ok(SuffixResult::Ambiguous(candidates)),
-    }
-}
-
-/// Escape LIKE metacharacters. All queries using this must include `ESCAPE '\'`.
-fn escape_like(input: &str) -> String {
-    input
-        .replace('\\', "\\\\")
-        .replace('%', "\\%")
-        .replace('_', "\\_")
-}
-
 // ─── Scoring Helpers ─────────────────────────────────────────────────────────
 
 /// Exponential half-life decay: `2^(-days / half_life)`.
@@ -1203,11 +978,10 @@ fn query_expert_details(
         .collect();
     let in_clause = placeholders.join(",");
 
-    let notes_indexed_by = "INDEXED BY idx_notes_diffnote_path_created";
     let sql = format!(
         "
         WITH signals AS (
-            -- 1. DiffNote reviewer
+            -- 1. DiffNote reviewer (matches both new_path and old_path for renamed files)
             SELECT
                 n.author_username AS username,
                 'reviewer' AS role,
@@ -1216,7 +990,7 @@ fn query_expert_details(
                 m.title AS title,
                 COUNT(*) AS note_count,
                 MAX(n.created_at) AS last_activity
-            FROM notes n {notes_indexed_by}
+            FROM notes n
             JOIN discussions d ON n.discussion_id = d.id
             JOIN merge_requests m ON d.merge_request_id = m.id
             JOIN projects p ON m.project_id = p.id
@@ -1224,8 +998,9 @@ fn query_expert_details(
               AND n.is_system = 0
               AND n.author_username IS NOT NULL
               AND (m.author_username IS NULL OR n.author_username != m.author_username)
-              AND m.state IN ('opened','merged')
-              AND n.position_new_path {path_op}
+              AND m.state IN ('opened','merged','closed')
+              AND (n.position_new_path {path_op}
+                   OR (n.position_old_path IS NOT NULL AND n.position_old_path {path_op}))
               AND n.created_at >= ?2
               AND (?3 IS NULL OR n.project_id = ?3)
               AND n.author_username IN ({in_clause})
@@ -1233,7 +1008,7 @@ fn query_expert_details(
 
             UNION ALL
 
-            -- 2. DiffNote MR author
+            -- 2. DiffNote MR author (matches both new_path and old_path for renamed files)
             SELECT
                 m.author_username AS username,
                 'author' AS role,
@@ -1244,13 +1019,14 @@ fn query_expert_details(
                 MAX(n.created_at) AS last_activity
             FROM merge_requests m
             JOIN discussions d ON d.merge_request_id = m.id
-            JOIN notes n {notes_indexed_by} ON n.discussion_id = d.id
+            JOIN notes n ON n.discussion_id = d.id
             JOIN projects p ON m.project_id = p.id
             WHERE n.note_type = 'DiffNote'
               AND n.is_system = 0
               AND m.author_username IS NOT NULL
-              AND m.state IN ('opened','merged')
-              AND n.position_new_path {path_op}
+              AND m.state IN ('opened','merged','closed')
+              AND (n.position_new_path {path_op}
+                   OR (n.position_old_path IS NOT NULL AND n.position_old_path {path_op}))
               AND n.created_at >= ?2
               AND (?3 IS NULL OR n.project_id = ?3)
               AND m.author_username IN ({in_clause})
@@ -1258,7 +1034,7 @@ fn query_expert_details(
 
             UNION ALL
 
-            -- 3. MR author via file changes
+            -- 3. MR author via file changes (matches both new_path and old_path)
             SELECT
                 m.author_username AS username,
                 'author' AS role,
@@ -1271,15 +1047,16 @@ fn query_expert_details(
             JOIN merge_requests m ON fc.merge_request_id = m.id
             JOIN projects p ON m.project_id = p.id
             WHERE m.author_username IS NOT NULL
-              AND m.state IN ('opened','merged')
-              AND fc.new_path {path_op}
+              AND m.state IN ('opened','merged','closed')
+              AND (fc.new_path {path_op}
+                   OR (fc.old_path IS NOT NULL AND fc.old_path {path_op}))
               AND m.updated_at >= ?2
               AND (?3 IS NULL OR fc.project_id = ?3)
               AND m.author_username IN ({in_clause})
 
             UNION ALL
 
-            -- 4. MR reviewer via file changes + mr_reviewers
+            -- 4. MR reviewer via file changes + mr_reviewers (matches both new_path and old_path)
             SELECT
                 r.username AS username,
                 'reviewer' AS role,
@@ -1294,8 +1071,9 @@ fn query_expert_details(
             JOIN mr_reviewers r ON r.merge_request_id = m.id
             WHERE r.username IS NOT NULL
               AND (m.author_username IS NULL OR r.username != m.author_username)
-              AND m.state IN ('opened','merged')
-              AND fc.new_path {path_op}
+              AND m.state IN ('opened','merged','closed')
+              AND (fc.new_path {path_op}
+                   OR (fc.old_path IS NOT NULL AND fc.old_path {path_op}))
               AND m.updated_at >= ?2
               AND (?3 IS NULL OR fc.project_id = ?3)
               AND r.username IN ({in_clause})
@@ -1874,50 +1652,51 @@ fn query_overlap(
     } else {
         "= ?1"
     };
-    // Force the partial index on DiffNote queries (same rationale as expert mode).
-    // Without this hint SQLite picks idx_notes_system (38% of rows) instead of
-    // idx_notes_diffnote_path_created (9.3% of rows): measured 50-133x slower.
-    let notes_indexed_by = "INDEXED BY idx_notes_diffnote_path_created";
+    // Match both new_path and old_path to capture activity on renamed files.
+    // INDEXED BY removed to allow OR across path columns; overlap runs once
+    // per command so the minor plan difference is acceptable.
     let sql = format!(
         "SELECT username, role, touch_count, last_seen_at, mr_refs FROM (
-            -- 1. DiffNote reviewer
+            -- 1. DiffNote reviewer (matches both new_path and old_path)
             SELECT
                 n.author_username AS username,
                 'reviewer' AS role,
                 COUNT(DISTINCT m.id) AS touch_count,
                 MAX(n.created_at) AS last_seen_at,
                 GROUP_CONCAT(DISTINCT (p.path_with_namespace || '!' || m.iid)) AS mr_refs
-            FROM notes n {notes_indexed_by}
+            FROM notes n
             JOIN discussions d ON n.discussion_id = d.id
             JOIN merge_requests m ON d.merge_request_id = m.id
             JOIN projects p ON m.project_id = p.id
             WHERE n.note_type = 'DiffNote'
-              AND n.position_new_path {path_op}
+              AND (n.position_new_path {path_op}
+                   OR (n.position_old_path IS NOT NULL AND n.position_old_path {path_op}))
               AND n.is_system = 0
               AND n.author_username IS NOT NULL
               AND (m.author_username IS NULL OR n.author_username != m.author_username)
-              AND m.state IN ('opened','merged')
+              AND m.state IN ('opened','merged','closed')
               AND n.created_at >= ?2
               AND (?3 IS NULL OR n.project_id = ?3)
             GROUP BY n.author_username
 
             UNION ALL
 
-            -- 2. DiffNote MR author
+            -- 2. DiffNote MR author (matches both new_path and old_path)
             SELECT
                 m.author_username AS username,
                 'author' AS role,
                 COUNT(DISTINCT m.id) AS touch_count,
                 MAX(n.created_at) AS last_seen_at,
                 GROUP_CONCAT(DISTINCT (p.path_with_namespace || '!' || m.iid)) AS mr_refs
-            FROM notes n {notes_indexed_by}
+            FROM notes n
             JOIN discussions d ON n.discussion_id = d.id
             JOIN merge_requests m ON d.merge_request_id = m.id
             JOIN projects p ON m.project_id = p.id
             WHERE n.note_type = 'DiffNote'
-              AND n.position_new_path {path_op}
+              AND (n.position_new_path {path_op}
+                   OR (n.position_old_path IS NOT NULL AND n.position_old_path {path_op}))
               AND n.is_system = 0
-              AND m.state IN ('opened', 'merged')
+              AND m.state IN ('opened','merged','closed')
               AND m.author_username IS NOT NULL
               AND n.created_at >= ?2
               AND (?3 IS NULL OR n.project_id = ?3)
@@ -1925,7 +1704,7 @@ fn query_overlap(
 
             UNION ALL
 
-            -- 3. MR author via file changes
+            -- 3. MR author via file changes (matches both new_path and old_path)
             SELECT
                 m.author_username AS username,
                 'author' AS role,
@@ -1936,15 +1715,16 @@ fn query_overlap(
             JOIN merge_requests m ON fc.merge_request_id = m.id
             JOIN projects p ON m.project_id = p.id
             WHERE m.author_username IS NOT NULL
-              AND m.state IN ('opened','merged')
-              AND fc.new_path {path_op}
+              AND m.state IN ('opened','merged','closed')
+              AND (fc.new_path {path_op}
+                   OR (fc.old_path IS NOT NULL AND fc.old_path {path_op}))
               AND m.updated_at >= ?2
               AND (?3 IS NULL OR fc.project_id = ?3)
             GROUP BY m.author_username
 
             UNION ALL
 
-            -- 4. MR reviewer via file changes + mr_reviewers
+            -- 4. MR reviewer via file changes + mr_reviewers (matches both new_path and old_path)
             SELECT
                 r.username AS username,
                 'reviewer' AS role,
@@ -1957,8 +1737,9 @@ fn query_overlap(
             JOIN mr_reviewers r ON r.merge_request_id = m.id
             WHERE r.username IS NOT NULL
               AND (m.author_username IS NULL OR r.username != m.author_username)
-              AND m.state IN ('opened','merged')
-              AND fc.new_path {path_op}
+              AND m.state IN ('opened','merged','closed')
+              AND (fc.new_path {path_op}
+                   OR (fc.old_path IS NOT NULL AND fc.old_path {path_op}))
               AND m.updated_at >= ?2
               AND (?3 IS NULL OR fc.project_id = ?3)
             GROUP BY r.username

src/core/mod.rs

@@ -9,6 +9,7 @@ pub mod lock;
 pub mod logging;
 pub mod metrics;
 pub mod note_parser;
+pub mod path_resolver;
 pub mod paths;
 pub mod payloads;
 pub mod project;

src/core/path_resolver.rs

@@ -0,0 +1,244 @@
+use rusqlite::Connection;
+
+use super::error::{LoreError, Result};
+
+// ─── SQL Helpers ─────────────────────────────────────────────────────────────
+
+/// Escape LIKE metacharacters (`%`, `_`, `\`).
+/// All queries using this must include `ESCAPE '\'`.
+pub fn escape_like(input: &str) -> String {
+    input
+        .replace('\\', "\\\\")
+        .replace('%', "\\%")
+        .replace('_', "\\_")
+}
+
+/// Normalize user-supplied repo paths to match stored DiffNote / file-change paths.
+/// - trims whitespace
+/// - strips leading "./" and "/" (repo-relative paths)
+/// - converts '\' to '/' when no '/' present (Windows paste)
+/// - collapses repeated "//"
+pub fn normalize_repo_path(input: &str) -> String {
+    let mut s = input.trim().to_string();
+    // Windows backslash normalization (only when no forward slashes present)
+    if s.contains('\\') && !s.contains('/') {
+        s = s.replace('\\', "/");
+    }
+    // Strip leading ./
+    while s.starts_with("./") {
+        s = s[2..].to_string();
+    }
+    // Strip leading /
+    s = s.trim_start_matches('/').to_string();
+    // Collapse repeated //
+    while s.contains("//") {
+        s = s.replace("//", "/");
+    }
+    s
+}
+
+// ─── Path Query Resolution ──────────────────────────────────────────────────
+
+/// Describes how to match a user-supplied path in SQL.
+#[derive(Debug)]
+pub struct PathQuery {
+    /// The parameter value to bind.
+    pub value: String,
+    /// If true: use `LIKE value ESCAPE '\'`. If false: use `= value`.
+    pub is_prefix: bool,
+}
+
+/// Result of a suffix probe against the DB.
+pub enum SuffixResult {
+    /// Suffix probe was not attempted (conditions not met).
+    NotAttempted,
+    /// No paths matched the suffix.
+    NoMatch,
+    /// Exactly one distinct path matched — auto-resolve.
+    Unique(String),
+    /// Multiple distinct paths matched — user must disambiguate.
+    Ambiguous(Vec<String>),
+}
+
+/// Build a path query from a user-supplied path, with project-scoped DB probes.
+///
+/// Resolution strategy (in priority order):
+/// 1. Trailing `/` → directory prefix (LIKE `path/%`)
+/// 2. Exact match probe against notes + `mr_file_changes` → exact (= `path`)
+/// 3. Directory prefix probe → prefix (LIKE `path/%`)
+/// 4. Suffix probe for bare filenames → auto-resolve or ambiguity error
+/// 5. Heuristic fallback: `.` in last segment → file, else → directory prefix
+pub fn build_path_query(
+    conn: &Connection,
+    path: &str,
+    project_id: Option<i64>,
+) -> Result<PathQuery> {
+    let trimmed = path.trim_end_matches('/');
+    let last_segment = trimmed.rsplit('/').next().unwrap_or(trimmed);
+    let is_root = !trimmed.contains('/');
+    let forced_dir = path.ends_with('/');
+    // Heuristic is now only a fallback; probes decide first when ambiguous.
+    let looks_like_file = !forced_dir && (is_root || last_segment.contains('.'));
+
+    // Probe 1: exact file exists in DiffNotes OR mr_file_changes (project-scoped)
+    // Checks both new_path and old_path to support querying renamed files.
+    let exact_exists = conn
+        .query_row(
+            "SELECT 1 FROM notes INDEXED BY idx_notes_diffnote_path_created
+         WHERE note_type = 'DiffNote'
+           AND is_system = 0
+           AND (position_new_path = ?1 OR position_old_path = ?1)
+           AND (?2 IS NULL OR project_id = ?2)
+         LIMIT 1",
+            rusqlite::params![trimmed, project_id],
+            |_| Ok(()),
+        )
+        .is_ok()
+        || conn
+            .query_row(
+                "SELECT 1 FROM mr_file_changes
+             WHERE (new_path = ?1 OR old_path = ?1)
+               AND (?2 IS NULL OR project_id = ?2)
+             LIMIT 1",
+                rusqlite::params![trimmed, project_id],
+                |_| Ok(()),
+            )
+            .is_ok();
+
+    // Probe 2: directory prefix exists in DiffNotes OR mr_file_changes (project-scoped)
+    let prefix_exists = if !forced_dir && !exact_exists {
+        let escaped = escape_like(trimmed);
+        let pat = format!("{escaped}/%");
+        conn.query_row(
+            "SELECT 1 FROM notes INDEXED BY idx_notes_diffnote_path_created
+             WHERE note_type = 'DiffNote'
+               AND is_system = 0
+               AND (position_new_path LIKE ?1 ESCAPE '\\' OR position_old_path LIKE ?1 ESCAPE '\\')
+               AND (?2 IS NULL OR project_id = ?2)
+             LIMIT 1",
+            rusqlite::params![pat, project_id],
+            |_| Ok(()),
+        )
+        .is_ok()
+            || conn
+                .query_row(
+                    "SELECT 1 FROM mr_file_changes
+                 WHERE (new_path LIKE ?1 ESCAPE '\\' OR old_path LIKE ?1 ESCAPE '\\')
+                   AND (?2 IS NULL OR project_id = ?2)
+                 LIMIT 1",
+                    rusqlite::params![pat, project_id],
+                    |_| Ok(()),
+                )
+                .is_ok()
+    } else {
+        false
+    };
+
+    // Probe 3: suffix match — user typed a bare filename or partial path that
+    // doesn't exist as-is. Search for full paths ending with /input (or equal to input).
+    // This handles "login.rs" matching "src/auth/login.rs".
+    let suffix_resolved = if !forced_dir && !exact_exists && !prefix_exists && looks_like_file {
+        suffix_probe(conn, trimmed, project_id)?
+    } else {
+        SuffixResult::NotAttempted
+    };
+
+    match suffix_resolved {
+        SuffixResult::Unique(full_path) => Ok(PathQuery {
+            value: full_path,
+            is_prefix: false,
+        }),
+        SuffixResult::Ambiguous(candidates) => {
+            let list = candidates
+                .iter()
+                .map(|p| format!("  {p}"))
+                .collect::<Vec<_>>()
+                .join("\n");
+            Err(LoreError::Ambiguous(format!(
+                "'{trimmed}' matches multiple paths. Use the full path or -p to scope:\n{list}"
+            )))
+        }
+        SuffixResult::NotAttempted | SuffixResult::NoMatch => {
+            // Original logic: exact > prefix > heuristic
+            let is_file = if forced_dir {
+                false
+            } else if exact_exists {
+                true
+            } else if prefix_exists {
+                false
+            } else {
+                looks_like_file
+            };
+
+            if is_file {
+                Ok(PathQuery {
+                    value: trimmed.to_string(),
+                    is_prefix: false,
+                })
+            } else {
+                let escaped = escape_like(trimmed);
+                Ok(PathQuery {
+                    value: format!("{escaped}/%"),
+                    is_prefix: true,
+                })
+            }
+        }
+    }
+}
+
+/// Probe both notes and mr_file_changes for paths ending with the given suffix.
+/// Searches both new_path and old_path columns to support renamed file resolution.
+/// Returns up to 11 distinct candidates (enough to detect ambiguity + show a useful list).
+pub fn suffix_probe(
+    conn: &Connection,
+    suffix: &str,
+    project_id: Option<i64>,
+) -> Result<SuffixResult> {
+    let escaped = escape_like(suffix);
+    let suffix_pat = format!("%/{escaped}");
+
+    let mut stmt = conn.prepare_cached(
+        "SELECT DISTINCT full_path FROM (
+            SELECT position_new_path AS full_path
+            FROM notes INDEXED BY idx_notes_diffnote_path_created
+            WHERE note_type = 'DiffNote'
+              AND is_system = 0
+              AND (position_new_path LIKE ?1 ESCAPE '\\' OR position_new_path = ?2)
+              AND (?3 IS NULL OR project_id = ?3)
+            UNION
+            SELECT new_path AS full_path FROM mr_file_changes
+            WHERE (new_path LIKE ?1 ESCAPE '\\' OR new_path = ?2)
+              AND (?3 IS NULL OR project_id = ?3)
+            UNION
+            SELECT position_old_path AS full_path FROM notes
+            WHERE note_type = 'DiffNote'
+              AND is_system = 0
+              AND position_old_path IS NOT NULL
+              AND (position_old_path LIKE ?1 ESCAPE '\\' OR position_old_path = ?2)
+              AND (?3 IS NULL OR project_id = ?3)
+            UNION
+            SELECT old_path AS full_path FROM mr_file_changes
+            WHERE old_path IS NOT NULL
+              AND (old_path LIKE ?1 ESCAPE '\\' OR old_path = ?2)
+              AND (?3 IS NULL OR project_id = ?3)
+        )
+        ORDER BY full_path
+        LIMIT 11",
+    )?;
+
+    let candidates: Vec<String> = stmt
+        .query_map(rusqlite::params![suffix_pat, suffix, project_id], |row| {
+            row.get(0)
+        })?
+        .collect::<std::result::Result<Vec<_>, _>>()?;
+
+    match candidates.len() {
+        0 => Ok(SuffixResult::NoMatch),
+        1 => Ok(SuffixResult::Unique(candidates.into_iter().next().unwrap())),
+        _ => Ok(SuffixResult::Ambiguous(candidates)),
+    }
+}
+
+#[cfg(test)]
+#[path = "path_resolver_tests.rs"]
+mod tests;

src/core/path_resolver_tests.rs

@@ -0,0 +1,290 @@
+use super::*;
+use crate::core::db::{create_connection, run_migrations};
+use std::path::Path;
+
+fn setup_test_db() -> Connection {
+    let conn = create_connection(Path::new(":memory:")).unwrap();
+    run_migrations(&conn).unwrap();
+    conn
+}
+
+fn seed_project(conn: &Connection, id: i64) {
+    conn.execute(
+        "INSERT INTO projects (id, gitlab_project_id, path_with_namespace, web_url, created_at, updated_at)
+         VALUES (?1, ?1, 'group/repo', 'https://gl.example.com/group/repo', 1000, 2000)",
+        rusqlite::params![id],
+    )
+    .unwrap();
+}
+
+fn seed_mr(conn: &Connection, mr_id: i64, project_id: i64) {
+    conn.execute(
+        "INSERT INTO merge_requests (id, gitlab_id, iid, project_id, title, state, \
+         created_at, updated_at, last_seen_at, source_branch, target_branch)
+         VALUES (?1, ?1, ?1, ?2, 'MR', 'merged', 1000, 2000, 2000, 'feat', 'main')",
+        rusqlite::params![mr_id, project_id],
+    )
+    .unwrap();
+}
+
+fn seed_file_change(conn: &Connection, mr_id: i64, project_id: i64, path: &str) {
+    conn.execute(
+        "INSERT INTO mr_file_changes (merge_request_id, project_id, new_path, change_type)
+         VALUES (?1, ?2, ?3, 'modified')",
+        rusqlite::params![mr_id, project_id, path],
+    )
+    .unwrap();
+}
+
+fn seed_diffnote(conn: &Connection, id: i64, project_id: i64, path: &str) {
+    // Need a discussion first (MergeRequest type, linked to mr_id=1)
+    conn.execute(
+        "INSERT OR IGNORE INTO discussions (id, gitlab_discussion_id, project_id, \
+         merge_request_id, noteable_type, resolvable, resolved, last_seen_at, last_note_at)
+         VALUES (?1, ?2, ?3, 1, 'MergeRequest', 1, 0, 2000, 2000)",
+        rusqlite::params![id, format!("disc-{id}"), project_id],
+    )
+    .unwrap();
+    conn.execute(
+        "INSERT INTO notes (id, gitlab_id, discussion_id, project_id, note_type, is_system, \
+         author_username, body, created_at, updated_at, last_seen_at, position_new_path)
+         VALUES (?1, ?1, ?1, ?2, 'DiffNote', 0, 'user', 'note', 1000, 2000, 2000, ?3)",
+        rusqlite::params![id, project_id, path],
+    )
+    .unwrap();
+}
+
+// ─── escape_like ─────────────────────────────────────────────────────────────
+
+#[test]
+fn test_escape_like() {
+    assert_eq!(escape_like("normal/path"), "normal/path");
+    assert_eq!(escape_like("has_underscore"), "has\\_underscore");
+    assert_eq!(escape_like("has%percent"), "has\\%percent");
+    assert_eq!(escape_like("has\\backslash"), "has\\\\backslash");
+}
+
+// ─── normalize_repo_path ─────────────────────────────────────────────────────
+
+#[test]
+fn test_normalize_repo_path() {
+    assert_eq!(normalize_repo_path("./src/foo/"), "src/foo/");
+    assert_eq!(normalize_repo_path("/src/foo/"), "src/foo/");
+    assert_eq!(normalize_repo_path("././src/foo"), "src/foo");
+    assert_eq!(normalize_repo_path("src\\foo\\bar.rs"), "src/foo/bar.rs");
+    assert_eq!(normalize_repo_path("src/foo\\bar"), "src/foo\\bar");
+    assert_eq!(normalize_repo_path("src//foo//bar/"), "src/foo/bar/");
+    assert_eq!(normalize_repo_path("  src/foo/  "), "src/foo/");
+    assert_eq!(normalize_repo_path("src/foo/bar.rs"), "src/foo/bar.rs");
+    assert_eq!(normalize_repo_path(""), "");
+}
+
+// ─── build_path_query heuristics (no DB data) ──────────────────────────────
+
+#[test]
+fn test_trailing_slash_is_prefix() {
+    let conn = setup_test_db();
+    let pq = build_path_query(&conn, "src/auth/", None).unwrap();
+    assert_eq!(pq.value, "src/auth/%");
+    assert!(pq.is_prefix);
+}
+
+#[test]
+fn test_no_dot_in_last_segment_is_prefix() {
+    let conn = setup_test_db();
+    let pq = build_path_query(&conn, "src/auth", None).unwrap();
+    assert_eq!(pq.value, "src/auth/%");
+    assert!(pq.is_prefix);
+}
+
+#[test]
+fn test_file_extension_is_exact() {
+    let conn = setup_test_db();
+    let pq = build_path_query(&conn, "src/auth/login.rs", None).unwrap();
+    assert_eq!(pq.value, "src/auth/login.rs");
+    assert!(!pq.is_prefix);
+}
+
+#[test]
+fn test_root_file_is_exact() {
+    let conn = setup_test_db();
+    let pq = build_path_query(&conn, "README.md", None).unwrap();
+    assert_eq!(pq.value, "README.md");
+    assert!(!pq.is_prefix);
+}
+
+#[test]
+fn test_dotless_root_file_is_exact() {
+    let conn = setup_test_db();
+    let pq = build_path_query(&conn, "Makefile", None).unwrap();
+    assert_eq!(pq.value, "Makefile");
+    assert!(!pq.is_prefix);
+
+    let pq = build_path_query(&conn, "LICENSE", None).unwrap();
+    assert_eq!(pq.value, "LICENSE");
+    assert!(!pq.is_prefix);
+}
+
+#[test]
+fn test_metacharacters_escaped_in_prefix() {
+    let conn = setup_test_db();
+    let pq = build_path_query(&conn, "src/test_files/", None).unwrap();
+    assert_eq!(pq.value, "src/test\\_files/%");
+    assert!(pq.is_prefix);
+}
+
+#[test]
+fn test_exact_value_not_escaped() {
+    let conn = setup_test_db();
+    let pq = build_path_query(&conn, "README_with_underscore.md", None).unwrap();
+    assert_eq!(pq.value, "README_with_underscore.md");
+    assert!(!pq.is_prefix);
+}
+
+// ─── build_path_query DB probes ─────────────────────────────────────────────
+
+#[test]
+fn test_db_probe_detects_dotless_file() {
+    // "src/Dockerfile" has no dot in last segment -> normally prefix.
+    // DB probe detects it's actually a file.
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_mr(&conn, 1, 1);
+    seed_diffnote(&conn, 1, 1, "src/Dockerfile");
+
+    let pq = build_path_query(&conn, "src/Dockerfile", None).unwrap();
+    assert_eq!(pq.value, "src/Dockerfile");
+    assert!(!pq.is_prefix);
+
+    // Without DB data -> falls through to prefix
+    let empty = setup_test_db();
+    let pq2 = build_path_query(&empty, "src/Dockerfile", None).unwrap();
+    assert!(pq2.is_prefix);
+}
+
+#[test]
+fn test_db_probe_via_file_changes() {
+    // Exact match via mr_file_changes even without notes
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_mr(&conn, 1, 1);
+    seed_file_change(&conn, 1, 1, "src/Dockerfile");
+
+    let pq = build_path_query(&conn, "src/Dockerfile", None).unwrap();
+    assert_eq!(pq.value, "src/Dockerfile");
+    assert!(!pq.is_prefix);
+}
+
+#[test]
+fn test_db_probe_project_scoped() {
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_project(&conn, 2);
+    seed_mr(&conn, 1, 1);
+    seed_diffnote(&conn, 1, 1, "infra/Makefile");
+
+    // Unscoped: finds it
+    assert!(
+        !build_path_query(&conn, "infra/Makefile", None)
+            .unwrap()
+            .is_prefix
+    );
+    // Scoped to project 1: finds it
+    assert!(
+        !build_path_query(&conn, "infra/Makefile", Some(1))
+            .unwrap()
+            .is_prefix
+    );
+    // Scoped to project 2: no data -> prefix
+    assert!(
+        build_path_query(&conn, "infra/Makefile", Some(2))
+            .unwrap()
+            .is_prefix
+    );
+}
+
+// ─── suffix resolution ──────────────────────────────────────────────────────
+
+#[test]
+fn test_suffix_resolves_bare_filename() {
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_mr(&conn, 1, 1);
+    seed_file_change(&conn, 1, 1, "src/auth/login.rs");
+
+    let pq = build_path_query(&conn, "login.rs", None).unwrap();
+    assert_eq!(pq.value, "src/auth/login.rs");
+    assert!(!pq.is_prefix);
+}
+
+#[test]
+fn test_suffix_resolves_partial_path() {
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_mr(&conn, 1, 1);
+    seed_file_change(&conn, 1, 1, "src/auth/login.rs");
+
+    let pq = build_path_query(&conn, "auth/login.rs", None).unwrap();
+    assert_eq!(pq.value, "src/auth/login.rs");
+    assert!(!pq.is_prefix);
+}
+
+#[test]
+fn test_suffix_ambiguous_returns_error() {
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_mr(&conn, 1, 1);
+    seed_file_change(&conn, 1, 1, "src/auth/utils.rs");
+    seed_file_change(&conn, 1, 1, "src/db/utils.rs");
+
+    let err = build_path_query(&conn, "utils.rs", None).unwrap_err();
+    let msg = err.to_string();
+    assert!(msg.contains("src/auth/utils.rs"), "candidates: {msg}");
+    assert!(msg.contains("src/db/utils.rs"), "candidates: {msg}");
+}
+
+#[test]
+fn test_suffix_scoped_to_project() {
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_project(&conn, 2);
+    seed_mr(&conn, 1, 1);
+    seed_mr(&conn, 2, 2);
+    seed_file_change(&conn, 1, 1, "src/utils.rs");
+    seed_file_change(&conn, 2, 2, "lib/utils.rs");
+
+    // Unscoped: ambiguous
+    assert!(build_path_query(&conn, "utils.rs", None).is_err());
+
+    // Scoped to project 1: resolves
+    let pq = build_path_query(&conn, "utils.rs", Some(1)).unwrap();
+    assert_eq!(pq.value, "src/utils.rs");
+}
+
+#[test]
+fn test_suffix_deduplicates_across_sources() {
+    // Same path in notes AND file_changes -> single match, not ambiguous
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_mr(&conn, 1, 1);
+    seed_file_change(&conn, 1, 1, "src/auth/login.rs");
+    seed_diffnote(&conn, 1, 1, "src/auth/login.rs");
+
+    let pq = build_path_query(&conn, "login.rs", None).unwrap();
+    assert_eq!(pq.value, "src/auth/login.rs");
+    assert!(!pq.is_prefix);
+}
+
+#[test]
+fn test_exact_match_preferred_over_suffix() {
+    let conn = setup_test_db();
+    seed_project(&conn, 1);
+    seed_mr(&conn, 1, 1);
+    seed_file_change(&conn, 1, 1, "README.md");
+    seed_file_change(&conn, 1, 1, "docs/README.md");
+
+    // "README.md" exists as exact match -> no ambiguity
+    let pq = build_path_query(&conn, "README.md", None).unwrap();
+    assert_eq!(pq.value, "README.md");
+    assert!(!pq.is_prefix);
+}

src/core/project.rs

@@ -1,6 +1,7 @@
 use rusqlite::Connection;
 
 use super::error::{LoreError, Result};
+use super::path_resolver::escape_like;
 
 pub fn resolve_project(conn: &Connection, project_str: &str) -> Result<i64> {
     let exact = conn.query_row(
@@ -106,13 +107,6 @@ pub fn resolve_project(conn: &Connection, project_str: &str) -> Result<i64> {
 
 /// Escape LIKE metacharacters so `%` and `_` in user input are treated as
 /// literals.  All queries using this must include `ESCAPE '\'`.
-fn escape_like(input: &str) -> String {
-    input
-        .replace('\\', "\\\\")
-        .replace('%', "\\%")
-        .replace('_', "\\_")
-}
-
 #[cfg(test)]
 #[path = "project_tests.rs"]
 mod tests;

src/core/timeline.rs

@@ -118,6 +118,8 @@ pub struct UnresolvedRef {
 #[derive(Debug, Clone, Serialize)]
 pub struct TimelineResult {
     pub query: String,
+    /// The search mode actually used for seeding (e.g. "hybrid", "lexical", "lexical (hybrid fallback)").
+    pub search_mode: String,
     pub events: Vec<TimelineEvent>,
     /// Total events before the `--limit` was applied (for meta.total_events vs meta.showing).
     #[serde(skip)]

src/core/timeline_seed.rs

@@ -5,23 +5,28 @@ use tracing::debug;
 
 use crate::core::error::Result;
 use crate::core::timeline::{EntityRef, TimelineEvent, TimelineEventType, resolve_entity_ref};
-use crate::search::{FtsQueryMode, to_fts_query};
+use crate::embedding::ollama::OllamaClient;
+use crate::search::{FtsQueryMode, SearchFilters, SearchMode, search_hybrid, to_fts_query};
 
 /// Result of the seed + hydrate phases.
 pub struct SeedResult {
     pub seed_entities: Vec<EntityRef>,
     pub evidence_notes: Vec<TimelineEvent>,
+    /// The search mode actually used (hybrid with fallback info).
+    pub search_mode: String,
 }
 
 /// Run the SEED + HYDRATE phases of the timeline pipeline.
 ///
-/// 1. SEED: FTS5 keyword search over documents -> matched document IDs
+/// 1. SEED: Hybrid search (FTS + vector via RRF) over documents -> matched document IDs
 /// 2. HYDRATE: Map document IDs -> source entities + top matched notes as evidence
 ///
+/// When `client` is `None` or Ollama is unavailable, falls back to FTS-only search.
 /// Discussion documents are resolved to their parent entity (issue or MR).
 /// Entities are deduplicated. Evidence notes are capped at `max_evidence`.
-pub fn seed_timeline(
+pub async fn seed_timeline(
     conn: &Connection,
+    client: Option<&OllamaClient>,
     query: &str,
     project_id: Option<i64>,
     since_ms: Option<i64>,
@@ -33,57 +38,110 @@ pub fn seed_timeline(
         return Ok(SeedResult {
             seed_entities: Vec::new(),
             evidence_notes: Vec::new(),
+            search_mode: "lexical".to_owned(),
         });
     }
 
-    let seed_entities = find_seed_entities(conn, &fts_query, project_id, since_ms, max_seeds)?;
+    // 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 {
+        project_id,
+        updated_since: since_ms,
+        limit: max_seeds.saturating_mul(3),
+        ..SearchFilters::default()
+    };
+
+    let (hybrid_results, warnings) = search_hybrid(
+        conn,
+        client,
+        query,
+        SearchMode::Hybrid,
+        &filters,
+        FtsQueryMode::Safe,
+    )
+    .await?;
+
+    let search_mode = if warnings
+        .iter()
+        .any(|w| w.contains("falling back") || w.contains("FTS only"))
+    {
+        "lexical (hybrid fallback)".to_owned()
+    } else if client.is_some() && !hybrid_results.is_empty() {
+        "hybrid".to_owned()
+    } else {
+        "lexical".to_owned()
+    };
+
+    for w in &warnings {
+        debug!(warning = %w, "hybrid search warning during timeline seeding");
+    }
+
+    let seed_entities = resolve_documents_to_entities(
+        conn,
+        &hybrid_results
+            .iter()
+            .map(|r| r.document_id)
+            .collect::<Vec<_>>(),
+        max_seeds,
+    )?;
+
+    // Evidence notes stay FTS-only (supplementary context, not worth a second embedding call)
     let evidence_notes = find_evidence_notes(conn, &fts_query, project_id, since_ms, max_evidence)?;
 
     Ok(SeedResult {
         seed_entities,
         evidence_notes,
+        search_mode,
     })
 }
 
-/// Find seed entities via FTS5 search, resolving discussions to their parent entity.
-fn find_seed_entities(
+/// Resolve a list of document IDs to deduplicated entity refs.
+/// Discussion documents are resolved to their parent entity (issue or MR).
+fn resolve_documents_to_entities(
     conn: &Connection,
-    fts_query: &str,
-    project_id: Option<i64>,
-    since_ms: Option<i64>,
-    max_seeds: usize,
+    document_ids: &[i64],
+    max_entities: usize,
 ) -> Result<Vec<EntityRef>> {
-    let sql = r"
+    if document_ids.is_empty() {
+        return Ok(Vec::new());
+    }
+
+    let placeholders: String = document_ids
+        .iter()
+        .map(|_| "?")
+        .collect::<Vec<_>>()
+        .join(",");
+    let sql = format!(
+        r"
         SELECT d.source_type, d.source_id, d.project_id,
                disc.issue_id, disc.merge_request_id
-        FROM documents_fts
-        JOIN documents d ON d.id = documents_fts.rowid
+        FROM documents d
         LEFT JOIN discussions disc ON disc.id = d.source_id AND d.source_type = 'discussion'
-        WHERE documents_fts MATCH ?1
-          AND (?2 IS NULL OR d.project_id = ?2)
-          AND (?3 IS NULL OR d.updated_at >= ?3)
-        ORDER BY rank
-        LIMIT ?4
-    ";
+        WHERE d.id IN ({placeholders})
+        ORDER BY CASE d.id {order_clause} END
+        ",
+        order_clause = document_ids
+            .iter()
+            .enumerate()
+            .map(|(i, id)| format!("WHEN {id} THEN {i}"))
+            .collect::<Vec<_>>()
+            .join(" "),
+    );
 
-    let mut stmt = conn.prepare(sql)?;
-    let rows = stmt.query_map(
-        rusqlite::params![
-            fts_query,
-            project_id,
-            since_ms,
-            max_seeds.saturating_mul(3) as i64
-        ],
-        |row| {
-            Ok((
-                row.get::<_, String>(0)?,
-                row.get::<_, i64>(1)?,
-                row.get::<_, i64>(2)?,
-                row.get::<_, Option<i64>>(3)?,
-                row.get::<_, Option<i64>>(4)?,
-            ))
-        },
-    )?;
+    let mut stmt = conn.prepare(&sql)?;
+    let params: Vec<&dyn rusqlite::types::ToSql> = document_ids
+        .iter()
+        .map(|id| id as &dyn rusqlite::types::ToSql)
+        .collect();
+    let rows = stmt.query_map(params.as_slice(), |row| {
+        Ok((
+            row.get::<_, String>(0)?,
+            row.get::<_, i64>(1)?,
+            row.get::<_, i64>(2)?,
+            row.get::<_, Option<i64>>(3)?,
+            row.get::<_, Option<i64>>(4)?,
+        ))
+    })?;
 
     let mut seen = HashSet::new();
     let mut entities = Vec::new();
@@ -116,7 +174,7 @@ fn find_seed_entities(
             entities.push(entity_ref);
         }
 
-        if entities.len() >= max_seeds {
+        if entities.len() >= max_entities {
             break;
         }
     }

src/core/timeline_seed_tests.rs

@@ -85,16 +85,18 @@ fn insert_note(
     conn.last_insert_rowid()
 }
 
-#[test]
-fn test_seed_empty_query_returns_empty() {
+#[tokio::test]
+async fn test_seed_empty_query_returns_empty() {
     let conn = setup_test_db();
-    let result = seed_timeline(&conn, "", None, None, 50, 10).unwrap();
+    let result = seed_timeline(&conn, None, "", None, None, 50, 10)
+        .await
+        .unwrap();
     assert!(result.seed_entities.is_empty());
     assert!(result.evidence_notes.is_empty());
 }
 
-#[test]
-fn test_seed_no_matches_returns_empty() {
+#[tokio::test]
+async fn test_seed_no_matches_returns_empty() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
     let issue_id = insert_test_issue(&conn, project_id, 1);
@@ -106,12 +108,14 @@ fn test_seed_no_matches_returns_empty() {
         "unrelated content here",
     );
 
-    let result = seed_timeline(&conn, "nonexistent_xyzzy_query", None, None, 50, 10).unwrap();
+    let result = seed_timeline(&conn, None, "nonexistent_xyzzy_query", None, None, 50, 10)
+        .await
+        .unwrap();
     assert!(result.seed_entities.is_empty());
 }
 
-#[test]
-fn test_seed_finds_issue() {
+#[tokio::test]
+async fn test_seed_finds_issue() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
     let issue_id = insert_test_issue(&conn, project_id, 42);
@@ -123,15 +127,17 @@ fn test_seed_finds_issue() {
         "authentication error in login flow",
     );
 
-    let result = seed_timeline(&conn, "authentication", None, None, 50, 10).unwrap();
+    let result = seed_timeline(&conn, None, "authentication", None, None, 50, 10)
+        .await
+        .unwrap();
     assert_eq!(result.seed_entities.len(), 1);
     assert_eq!(result.seed_entities[0].entity_type, "issue");
     assert_eq!(result.seed_entities[0].entity_iid, 42);
     assert_eq!(result.seed_entities[0].project_path, "group/project");
 }
 
-#[test]
-fn test_seed_finds_mr() {
+#[tokio::test]
+async fn test_seed_finds_mr() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
     let mr_id = insert_test_mr(&conn, project_id, 99);
@@ -143,14 +149,16 @@ fn test_seed_finds_mr() {
         "fix authentication bug",
     );
 
-    let result = seed_timeline(&conn, "authentication", None, None, 50, 10).unwrap();
+    let result = seed_timeline(&conn, None, "authentication", None, None, 50, 10)
+        .await
+        .unwrap();
     assert_eq!(result.seed_entities.len(), 1);
     assert_eq!(result.seed_entities[0].entity_type, "merge_request");
     assert_eq!(result.seed_entities[0].entity_iid, 99);
 }
 
-#[test]
-fn test_seed_deduplicates_entities() {
+#[tokio::test]
+async fn test_seed_deduplicates_entities() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
     let issue_id = insert_test_issue(&conn, project_id, 10);
@@ -172,14 +180,16 @@ fn test_seed_deduplicates_entities() {
         "authentication error second doc",
     );
 
-    let result = seed_timeline(&conn, "authentication", None, None, 50, 10).unwrap();
+    let result = seed_timeline(&conn, None, "authentication", None, None, 50, 10)
+        .await
+        .unwrap();
     // Should deduplicate: both map to the same issue
     assert_eq!(result.seed_entities.len(), 1);
     assert_eq!(result.seed_entities[0].entity_iid, 10);
 }
 
-#[test]
-fn test_seed_resolves_discussion_to_parent() {
+#[tokio::test]
+async fn test_seed_resolves_discussion_to_parent() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
     let issue_id = insert_test_issue(&conn, project_id, 7);
@@ -192,14 +202,16 @@ fn test_seed_resolves_discussion_to_parent() {
         "deployment pipeline failed",
     );
 
-    let result = seed_timeline(&conn, "deployment", None, None, 50, 10).unwrap();
+    let result = seed_timeline(&conn, None, "deployment", None, None, 50, 10)
+        .await
+        .unwrap();
     assert_eq!(result.seed_entities.len(), 1);
     assert_eq!(result.seed_entities[0].entity_type, "issue");
     assert_eq!(result.seed_entities[0].entity_iid, 7);
 }
 
-#[test]
-fn test_seed_evidence_capped() {
+#[tokio::test]
+async fn test_seed_evidence_capped() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
     let issue_id = insert_test_issue(&conn, project_id, 1);
@@ -223,12 +235,14 @@ fn test_seed_evidence_capped() {
         );
     }
 
-    let result = seed_timeline(&conn, "deployment", None, None, 50, 5).unwrap();
+    let result = seed_timeline(&conn, None, "deployment", None, None, 50, 5)
+        .await
+        .unwrap();
     assert!(result.evidence_notes.len() <= 5);
 }
 
-#[test]
-fn test_seed_evidence_snippet_truncated() {
+#[tokio::test]
+async fn test_seed_evidence_snippet_truncated() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
     let issue_id = insert_test_issue(&conn, project_id, 1);
@@ -244,7 +258,9 @@ fn test_seed_evidence_snippet_truncated() {
     let long_body = "x".repeat(500);
     insert_note(&conn, disc_id, project_id, &long_body, false);
 
-    let result = seed_timeline(&conn, "deployment", None, None, 50, 10).unwrap();
+    let result = seed_timeline(&conn, None, "deployment", None, None, 50, 10)
+        .await
+        .unwrap();
     assert!(!result.evidence_notes.is_empty());
     if let TimelineEventType::NoteEvidence { snippet, .. } = &result.evidence_notes[0].event_type {
         assert!(snippet.chars().count() <= 200);
@@ -253,8 +269,8 @@ fn test_seed_evidence_snippet_truncated() {
     }
 }
 
-#[test]
-fn test_seed_respects_project_filter() {
+#[tokio::test]
+async fn test_seed_respects_project_filter() {
     let conn = setup_test_db();
     let project_id = insert_test_project(&conn);
 
@@ -285,7 +301,17 @@ fn test_seed_respects_project_filter() {
     );
 
     // Filter to project 1 only
-    let result = seed_timeline(&conn, "authentication", Some(project_id), None, 50, 10).unwrap();
+    let result = seed_timeline(
+        &conn,
+        None,
+        "authentication",
+        Some(project_id),
+        None,
+        50,
+        10,
+    )
+    .await
+    .unwrap();
     assert_eq!(result.seed_entities.len(), 1);
     assert_eq!(result.seed_entities[0].project_path, "group/project");
 }

src/main.rs

@@ -179,7 +179,9 @@ async fn main() {
         Some(Commands::Search(args)) => {
             handle_search(cli.config.as_deref(), args, robot_mode).await
         }
-        Some(Commands::Timeline(args)) => handle_timeline(cli.config.as_deref(), args, robot_mode),
+        Some(Commands::Timeline(args)) => {
+            handle_timeline(cli.config.as_deref(), args, robot_mode).await
+        }
         Some(Commands::Who(args)) => handle_who(cli.config.as_deref(), args, robot_mode),
         Some(Commands::Drift {
             entity_type,
@@ -1763,7 +1765,7 @@ async fn handle_stats(
     Ok(())
 }
 
-fn handle_timeline(
+async fn handle_timeline(
     config_override: Option<&str>,
     args: TimelineArgs,
     robot_mode: bool,
@@ -1784,7 +1786,7 @@ fn handle_timeline(
         max_evidence: args.max_evidence,
     };
 
-    let result = run_timeline(&config, &params)?;
+    let result = run_timeline(&config, &params).await?;
 
     if robot_mode {
         print_timeline_json_with_meta(

src/search/filters.rs

@@ -1,4 +1,5 @@
 use crate::core::error::Result;
+use crate::core::path_resolver::escape_like;
 use crate::documents::SourceType;
 use rusqlite::Connection;
 
@@ -43,12 +44,6 @@ impl SearchFilters {
     }
 }
 
-fn escape_like(s: &str) -> String {
-    s.replace('\\', "\\\\")
-        .replace('%', "\\%")
-        .replace('_', "\\_")
-}
-
 pub fn apply_filters(
     conn: &Connection,
     document_ids: &[i64],

src/search/vector.rs

@@ -40,6 +40,17 @@ fn max_chunks_per_document(conn: &Connection) -> Result<i64> {
         .unwrap_or(1))
 }
 
+/// sqlite-vec hard limit for KNN `k` parameter.
+const SQLITE_VEC_KNN_MAX: usize = 4_096;
+
+/// Compute the KNN k value from the requested limit and the max chunks per
+/// document. The result is guaranteed to never exceed [`SQLITE_VEC_KNN_MAX`].
+fn compute_knn_k(limit: usize, max_chunks_per_doc: i64) -> usize {
+    let max_chunks = max_chunks_per_doc.unsigned_abs().max(1) as usize;
+    let multiplier = (max_chunks * 3 / 2 + 1).clamp(8, 200);
+    (limit * multiplier).min(SQLITE_VEC_KNN_MAX)
+}
+
 pub fn search_vector(
     conn: &Connection,
     query_embedding: &[f32],
@@ -55,8 +66,7 @@ pub fn search_vector(
         .collect();
 
     let max_chunks = max_chunks_per_document(conn)?.max(1);
-    let multiplier = ((max_chunks.unsigned_abs() as usize * 3 / 2) + 1).clamp(8, 200);
-    let k = (limit * multiplier).min(10_000);
+    let k = compute_knn_k(limit, max_chunks);
 
     let mut stmt = conn.prepare(
         "SELECT rowid, distance
@@ -124,6 +134,52 @@ mod tests {
         assert_eq!(results.len(), 2);
     }
 
+    #[test]
+    fn test_knn_k_never_exceeds_sqlite_vec_limit() {
+        for limit in [1, 10, 50, 100, 500, 1000, 1500, 2000, 5000] {
+            for max_chunks in [1, 2, 5, 10, 50, 100, 200, 500, 1000] {
+                let k = compute_knn_k(limit, max_chunks);
+                assert!(
+                    k <= SQLITE_VEC_KNN_MAX,
+                    "k={k} exceeded limit for limit={limit}, max_chunks={max_chunks}"
+                );
+            }
+        }
+    }
+
+    #[test]
+    fn test_knn_k_reproduces_original_bug_scenario() {
+        let k = compute_knn_k(1500, 1);
+        assert!(
+            k <= SQLITE_VEC_KNN_MAX,
+            "k={k} exceeded 4096 at RECALL_CAP with 1 chunk"
+        );
+    }
+
+    #[test]
+    fn test_knn_k_small_limit_uses_minimum_multiplier() {
+        let k = compute_knn_k(10, 1);
+        assert_eq!(k, 80);
+    }
+
+    #[test]
+    fn test_knn_k_high_chunks_caps_multiplier() {
+        let k = compute_knn_k(10, 200);
+        assert_eq!(k, 2000);
+    }
+
+    #[test]
+    fn test_knn_k_zero_max_chunks_treated_as_one() {
+        let k = compute_knn_k(10, 0);
+        assert_eq!(k, 80);
+    }
+
+    #[test]
+    fn test_knn_k_negative_max_chunks_uses_absolute() {
+        let k = compute_knn_k(10, -5);
+        assert_eq!(k, compute_knn_k(10, 5));
+    }
+
     fn search_vector_dedup(rows: Vec<(i64, f64)>, limit: usize) -> Vec<VectorResult> {
         let mut best: HashMap<i64, f64> = HashMap::new();
         for (rowid, distance) in rows {

tests/timeline_pipeline_tests.rs

@@ -108,8 +108,8 @@ fn insert_label_event(
 
 /// Full pipeline: seed -> expand -> collect for a scenario with an issue
 /// that has a closing MR, state changes, and label events.
-#[test]
-fn pipeline_seed_expand_collect_end_to_end() {
+#[tokio::test]
+async fn pipeline_seed_expand_collect_end_to_end() {
     let conn = setup_db();
     let project_id = insert_project(&conn, "group/project");
 
@@ -149,7 +149,9 @@ fn pipeline_seed_expand_collect_end_to_end() {
     insert_label_event(&conn, project_id, Some(issue_id), "bug", 1500);
 
     // SEED: find entities matching "authentication"
-    let seed_result = seed_timeline(&conn, "authentication", None, None, 50, 10).unwrap();
+    let seed_result = seed_timeline(&conn, None, "authentication", None, None, 50, 10)
+        .await
+        .unwrap();
     assert!(
         !seed_result.seed_entities.is_empty(),
         "Seed should find at least one entity"
@@ -213,12 +215,14 @@ fn pipeline_seed_expand_collect_end_to_end() {
 }
 
 /// Verify the pipeline handles an empty FTS result gracefully.
-#[test]
-fn pipeline_empty_query_produces_empty_result() {
+#[tokio::test]
+async fn pipeline_empty_query_produces_empty_result() {
     let conn = setup_db();
     let _project_id = insert_project(&conn, "group/project");
 
-    let seed_result = seed_timeline(&conn, "", None, None, 50, 10).unwrap();
+    let seed_result = seed_timeline(&conn, None, "", None, None, 50, 10)
+        .await
+        .unwrap();
     assert!(seed_result.seed_entities.is_empty());
 
     let expand_result = expand_timeline(&conn, &seed_result.seed_entities, 1, false, 100).unwrap();
@@ -237,8 +241,8 @@ fn pipeline_empty_query_produces_empty_result() {
 }
 
 /// Verify since filter propagates through the full pipeline.
-#[test]
-fn pipeline_since_filter_excludes_old_events() {
+#[tokio::test]
+async fn pipeline_since_filter_excludes_old_events() {
     let conn = setup_db();
     let project_id = insert_project(&conn, "group/project");
 
@@ -255,7 +259,9 @@ fn pipeline_since_filter_excludes_old_events() {
     insert_state_event(&conn, project_id, Some(issue_id), None, "closed", 2000);
     insert_state_event(&conn, project_id, Some(issue_id), None, "reopened", 8000);
 
-    let seed_result = seed_timeline(&conn, "deploy", None, None, 50, 10).unwrap();
+    let seed_result = seed_timeline(&conn, None, "deploy", None, None, 50, 10)
+        .await
+        .unwrap();
     let expand_result = expand_timeline(&conn, &seed_result.seed_entities, 0, false, 100).unwrap();
 
     // Collect with since=5000: should exclude Created(1000) and closed(2000)
@@ -274,8 +280,8 @@ fn pipeline_since_filter_excludes_old_events() {
 }
 
 /// Verify unresolved references use Option<i64> for target_iid.
-#[test]
-fn pipeline_unresolved_refs_have_optional_iid() {
+#[tokio::test]
+async fn pipeline_unresolved_refs_have_optional_iid() {
     let conn = setup_db();
     let project_id = insert_project(&conn, "group/project");
 
@@ -302,7 +308,9 @@ fn pipeline_unresolved_refs_have_optional_iid() {
     )
     .unwrap();
 
-    let seed_result = seed_timeline(&conn, "cross project", None, None, 50, 10).unwrap();
+    let seed_result = seed_timeline(&conn, None, "cross project", None, None, 50, 10)
+        .await
+        .unwrap();
     let expand_result = expand_timeline(&conn, &seed_result.seed_entities, 1, false, 100).unwrap();
 
     assert_eq!(expand_result.unresolved_references.len(), 2);