episteme 0.3.9

Knowledge graph for software engineering — design patterns, refactorings, and laws for AI agents
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
use std::collections::HashMap;
use std::sync::Mutex;

use rusqlite::{Connection, params};

use crate::domain::types::{CorrelationScore, GraphEdge, UserEntity};
use crate::ports::graph::MutableGraphRepository;

// ---------------------------------------------------------------------------
// Helper for neighbor queries (avoids closure type mismatch)
// ---------------------------------------------------------------------------

fn query_neighbors(
    conn: &Connection,
    base_sql: &str,
    entity_id: &str,
    relation_type: Option<&str>,
    out: &mut Vec<String>,
) {
    let sql = if relation_type.is_some() {
        format!("{base_sql} AND relation_type=?2")
    } else {
        base_sql.to_owned()
    };
    let Ok(mut stmt) = conn.prepare(&sql) else {
        return;
    };
    let rows: Vec<String> = if let Some(rt) = relation_type {
        stmt.query_map(params![entity_id, rt], |row| row.get::<_, String>(0))
            .ok()
            .map(|r| r.filter_map(|v| v.ok()).collect())
            .unwrap_or_default()
    } else {
        stmt.query_map(params![entity_id], |row| row.get::<_, String>(0))
            .ok()
            .map(|r| r.filter_map(|v| v.ok()).collect())
            .unwrap_or_default()
    };
    out.extend(rows);
}

// ---------------------------------------------------------------------------
// Schema initialization
// ---------------------------------------------------------------------------

fn init_schema(conn: &Connection) -> Result<(), String> {
    conn.execute_batch(
        "
        PRAGMA journal_mode=WAL;

        CREATE TABLE IF NOT EXISTS user_entities (
            id TEXT PRIMARY KEY,
            title TEXT NOT NULL,
            content TEXT NOT NULL,
            author TEXT NOT NULL DEFAULT 'user',
            confidence REAL NOT NULL DEFAULT 0.5,
            evidence_count INTEGER NOT NULL DEFAULT 0,
            last_validated TEXT NOT NULL DEFAULT '',
            tags TEXT NOT NULL DEFAULT '[]',
            relations TEXT NOT NULL DEFAULT '{}',
            created_at TEXT NOT NULL DEFAULT '',
            updated_at TEXT NOT NULL DEFAULT ''
        );

        CREATE TABLE IF NOT EXISTS user_relations (
            id INTEGER PRIMARY KEY AUTOINCREMENT,
            from_id TEXT NOT NULL,
            relation_type TEXT NOT NULL,
            to_id TEXT NOT NULL,
            UNIQUE(from_id, relation_type, to_id),
            FOREIGN KEY (from_id) REFERENCES user_entities(id) ON DELETE CASCADE
        );

        CREATE TABLE IF NOT EXISTS user_embeddings (
            entity_id TEXT PRIMARY KEY,
            embedding BLOB NOT NULL,
            FOREIGN KEY (entity_id) REFERENCES user_entities(id) ON DELETE CASCADE
        );

        CREATE VIRTUAL TABLE IF NOT EXISTS user_entities_fts USING fts5(
            title, content, tags, content=user_entities, content_rowid=rowid
        );

        CREATE INDEX IF NOT EXISTS idx_user_relations_from ON user_relations(from_id);
        CREATE INDEX IF NOT EXISTS idx_user_relations_to ON user_relations(to_id);

        -- Sequence counter for atomic TK-xxx ID generation
        CREATE TABLE IF NOT EXISTS insight_seq (key TEXT PRIMARY KEY, val INTEGER NOT NULL);
        INSERT OR IGNORE INTO insight_seq (key, val) VALUES ('tk', 0);

        -- FTS5 content-sync triggers (required for content= tables)
        CREATE TRIGGER IF NOT EXISTS user_entities_ai AFTER INSERT ON user_entities BEGIN
            INSERT INTO user_entities_fts(rowid, title, content, tags)
            VALUES (new.rowid, new.title, new.content, new.tags);
        END;

        CREATE TRIGGER IF NOT EXISTS user_entities_ad AFTER DELETE ON user_entities BEGIN
            INSERT INTO user_entities_fts(user_entities_fts, rowid, title, content, tags)
            VALUES ('delete', old.rowid, old.title, old.content, old.tags);
        END;

        CREATE TRIGGER IF NOT EXISTS user_entities_au AFTER UPDATE ON user_entities BEGIN
            INSERT INTO user_entities_fts(user_entities_fts, rowid, title, content, tags)
            VALUES ('delete', old.rowid, old.title, old.content, old.tags);
            INSERT INTO user_entities_fts(rowid, title, content, tags)
            VALUES (new.rowid, new.title, new.content, new.tags);
        END;
        ",
    )
    .map_err(|e| format!("user graph schema init: {e}"))?;

    // Migrations: add columns that may not exist in older databases
    if let Err(e) = conn.execute_batch(
        "ALTER TABLE user_entities ADD COLUMN link_provenance TEXT NOT NULL DEFAULT '{}'",
    ) {
        let msg = e.to_string();
        if !msg.contains("duplicate column") {
            return Err(format!("migration link_provenance: {msg}"));
        }
    }

    Ok(())
}

// ---------------------------------------------------------------------------
// UserGraphStore
// ---------------------------------------------------------------------------

pub struct UserGraphStore {
    conn: Mutex<Connection>,
}

impl UserGraphStore {
    pub fn open(path: &std::path::Path) -> Result<Self, String> {
        let conn = Connection::open(path).map_err(|e| format!("open user graph: {e}"))?;
        conn.execute_batch("PRAGMA foreign_keys=ON;")
            .map_err(|e| format!("enable FK: {e}"))?;
        init_schema(&conn)?;
        Ok(Self {
            conn: Mutex::new(conn),
        })
    }

    pub fn open_in_memory() -> Result<Self, String> {
        let conn = Connection::open_in_memory().map_err(|e| format!("open in-memory: {e}"))?;
        init_schema(&conn)?;
        Ok(Self {
            conn: Mutex::new(conn),
        })
    }

    /// Atomically generate the next TK-xxx ID using a SQLite sequence counter.
    /// Syncs the counter with any manually-added entities to prevent collisions.
    /// Safe under concurrent writes (MCP server + CLI simultaneously).
    pub fn next_insight_id_atomic(&self) -> Result<String, String> {
        let conn = self
            .conn
            .lock()
            .map_err(|e| format!("mutex poisoned: {e}"))?;

        // Sync: ensure sequence is at least as high as any existing TK-xxx entity
        let max_existing: i64 = conn
            .query_row(
                "SELECT COALESCE(MAX(CAST(SUBSTR(id, 4) AS INTEGER)), 0) \
                 FROM user_entities WHERE id LIKE 'TK-%'",
                [],
                |row| row.get::<_, i64>(0),
            )
            .unwrap_or(0);
        if max_existing > 0 {
            conn.execute(
                "UPDATE insight_seq SET val = MAX(val, ?1) WHERE key = 'tk'",
                params![max_existing],
            )
            .map_err(|e| format!("sync sequence: {e}"))?;
        }

        // Atomic increment and return
        let next: i64 = conn
            .query_row(
                "UPDATE insight_seq SET val = val + 1 WHERE key = 'tk' RETURNING val",
                [],
                |row| row.get(0),
            )
            .map_err(|e| format!("next insight id: {e}"))?;
        Ok(format!("TK-{:03}", next))
    }
}

// ---------------------------------------------------------------------------
// Row mapping helpers
// ---------------------------------------------------------------------------

fn row_to_user_entity(row: &rusqlite::Row<'_>) -> rusqlite::Result<UserEntity> {
    let tags_str: String = row.get(7)?;
    let rels_str: String = row.get(8)?;
    let prov_str: String = row.get(11)?;
    Ok(UserEntity {
        id: row.get(0)?,
        title: row.get(1)?,
        content: row.get(2)?,
        author: row.get(3)?,
        confidence: row.get(4)?,
        evidence_count: row.get(5)?,
        last_validated: row.get(6)?,
        tags: serde_json::from_str(&tags_str).unwrap_or_default(),
        relations: serde_json::from_str(&rels_str).unwrap_or_default(),
        link_provenance: serde_json::from_str(&prov_str).unwrap_or_default(),
        created_at: row.get(9)?,
        updated_at: row.get(10)?,
    })
}

// ---------------------------------------------------------------------------
// MutableGraphRepository implementation
// ---------------------------------------------------------------------------

impl MutableGraphRepository for UserGraphStore {
    fn add_entity(&self, entity: UserEntity) -> Result<(), String> {
        let conn = self
            .conn
            .lock()
            .map_err(|e| format!("mutex poisoned: {e}"))?;
        let tags = serde_json::to_string(&entity.tags).unwrap_or_else(|_| "[]".to_owned());
        let rels = serde_json::to_string(&entity.relations).unwrap_or_else(|_| "{}".to_owned());
        let prov =
            serde_json::to_string(&entity.link_provenance).unwrap_or_else(|_| "{}".to_owned());
        conn.execute(
            "INSERT INTO user_entities (id, title, content, author, confidence, evidence_count, last_validated, tags, relations, link_provenance, created_at, updated_at)
             VALUES (?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9, ?10, ?11, ?12)",
            params![
                entity.id, entity.title, entity.content, entity.author,
                entity.confidence, entity.evidence_count, entity.last_validated,
                tags, rels, prov, entity.created_at, entity.updated_at,
            ],
        )
        .map_err(|e| format!("insert user entity: {e}"))?;

        // Insert relations
        for (rel_type, targets) in &entity.relations {
            for target in targets {
                conn.execute(
                    "INSERT OR IGNORE INTO user_relations (from_id, relation_type, to_id) VALUES (?1, ?2, ?3)",
                    params![entity.id, rel_type, target],
                )
                .map_err(|e| format!("insert relation: {e}"))?;
            }
        }
        Ok(())
    }

    fn update_entity(&self, id: &str, entity: UserEntity) -> Result<(), String> {
        let conn = self
            .conn
            .lock()
            .map_err(|e| format!("mutex poisoned: {e}"))?;
        let tags = serde_json::to_string(&entity.tags).unwrap_or_else(|_| "[]".to_owned());
        let rels = serde_json::to_string(&entity.relations).unwrap_or_else(|_| "{}".to_owned());
        let prov =
            serde_json::to_string(&entity.link_provenance).unwrap_or_else(|_| "{}".to_owned());
        let rows = conn
            .execute(
                "UPDATE user_entities SET title=?1, content=?2, author=?3, confidence=?4,
                 evidence_count=?5, last_validated=?6, tags=?7, relations=?8, link_provenance=?9, updated_at=?10
                 WHERE id=?11",
                params![
                    entity.title,
                    entity.content,
                    entity.author,
                    entity.confidence,
                    entity.evidence_count,
                    entity.last_validated,
                    tags,
                    rels,
                    prov,
                    entity.updated_at,
                    id,
                ],
            )
            .map_err(|e| format!("update user entity: {e}"))?;
        if rows == 0 {
            return Err(format!("entity not found: {id}"));
        }

        // Re-sync relations: delete old, insert new
        conn.execute("DELETE FROM user_relations WHERE from_id=?1", params![id])
            .map_err(|e| format!("delete old relations: {e}"))?;
        for (rel_type, targets) in &entity.relations {
            for target in targets {
                conn.execute(
                    "INSERT OR IGNORE INTO user_relations (from_id, relation_type, to_id) VALUES (?1, ?2, ?3)",
                    params![id, rel_type, target],
                )
                .map_err(|e| format!("insert relation: {e}"))?;
            }
        }
        Ok(())
    }

    fn remove_entity(&self, id: &str) -> Result<(), String> {
        let conn = self
            .conn
            .lock()
            .map_err(|e| format!("mutex poisoned: {e}"))?;
        conn.execute(
            "DELETE FROM user_relations WHERE from_id=?1 OR to_id=?1",
            params![id],
        )
        .map_err(|e| format!("delete relations: {e}"))?;
        conn.execute(
            "DELETE FROM user_embeddings WHERE entity_id=?1",
            params![id],
        )
        .map_err(|e| format!("delete embedding: {e}"))?;
        let rows = conn
            .execute("DELETE FROM user_entities WHERE id=?1", params![id])
            .map_err(|e| format!("delete entity: {e}"))?;
        if rows == 0 {
            return Err(format!("entity not found: {id}"));
        }
        Ok(())
    }

    fn add_relation(&self, from: &str, relation: &str, to: &str) -> Result<(), String> {
        let conn = self
            .conn
            .lock()
            .map_err(|e| format!("mutex poisoned: {e}"))?;
        conn.execute(
            "INSERT OR IGNORE INTO user_relations (from_id, relation_type, to_id) VALUES (?1, ?2, ?3)",
            params![from, relation, to],
        )
        .map_err(|e| format!("insert relation: {e}"))?;
        Ok(())
    }

    fn remove_relation(&self, from: &str, relation: &str, to: &str) -> Result<(), String> {
        let conn = self
            .conn
            .lock()
            .map_err(|e| format!("mutex poisoned: {e}"))?;
        conn.execute(
            "DELETE FROM user_relations WHERE from_id=?1 AND relation_type=?2 AND to_id=?3",
            params![from, relation, to],
        )
        .map_err(|e| format!("delete relation: {e}"))?;
        Ok(())
    }

    fn get_user_entity(&self, id: &str) -> Option<UserEntity> {
        let conn = self.conn.lock().ok()?;
        let mut stmt = conn
            .prepare(
                "SELECT id, title, content, author, confidence, evidence_count, last_validated, tags, relations, created_at, updated_at, link_provenance FROM user_entities WHERE id=?1",
            )
            .ok()?;
        stmt.query_row(params![id], row_to_user_entity).ok()
    }

    fn all_user_entity_ids(&self) -> Vec<String> {
        let conn = match self.conn.lock() {
            Ok(c) => c,
            Err(_) => return Vec::new(),
        };
        let mut stmt = match conn.prepare("SELECT id FROM user_entities") {
            Ok(s) => s,
            Err(_) => return Vec::new(),
        };
        let rows = match stmt.query_map([], |row| row.get::<_, String>(0)) {
            Ok(r) => r,
            Err(_) => return Vec::new(),
        };
        rows.filter_map(|r| r.ok()).collect()
    }

    fn all_user_entities(&self) -> Vec<UserEntity> {
        let conn = match self.conn.lock() {
            Ok(c) => c,
            Err(_) => return Vec::new(),
        };
        let mut stmt = match conn.prepare(
            "SELECT id, title, content, author, confidence, evidence_count,
                    last_validated, tags, relations, created_at, updated_at, link_provenance
             FROM user_entities",
        ) {
            Ok(s) => s,
            Err(_) => return Vec::new(),
        };
        let rows = stmt.query_map([], row_to_user_entity);
        match rows {
            Ok(r) => r.filter_map(|x| x.ok()).collect(),
            Err(_) => Vec::new(),
        }
    }

    fn get_user_neighbors(&self, entity_id: &str, relation_type: Option<&str>) -> Vec<String> {
        let conn = match self.conn.lock() {
            Ok(c) => c,
            Err(_) => return Vec::new(),
        };
        let mut neighbors = Vec::new();

        // Outgoing
        query_neighbors(
            &conn,
            "SELECT to_id FROM user_relations WHERE from_id=?1",
            entity_id,
            relation_type,
            &mut neighbors,
        );
        // Incoming
        query_neighbors(
            &conn,
            "SELECT from_id FROM user_relations WHERE to_id=?1",
            entity_id,
            relation_type,
            &mut neighbors,
        );

        neighbors.sort();
        neighbors.dedup();
        neighbors
    }

    fn get_user_all_edges(&self, entity_id: &str) -> Vec<GraphEdge> {
        let conn = match self.conn.lock() {
            Ok(c) => c,
            Err(_) => return Vec::new(),
        };
        let mut edges = Vec::new();

        // Outgoing
        let mut stmt = match conn
            .prepare("SELECT from_id, relation_type, to_id FROM user_relations WHERE from_id=?1")
        {
            Ok(s) => s,
            Err(_) => return Vec::new(),
        };
        let rows = stmt.query_map(params![entity_id], |row| {
            Ok(GraphEdge {
                from_id: row.get(0)?,
                relation_type: row.get(1)?,
                to_id: row.get(2)?,
            })
        });
        if let Ok(rows) = rows {
            for r in rows.flatten() {
                edges.push(r);
            }
        }

        // Incoming
        let mut stmt = match conn
            .prepare("SELECT from_id, relation_type, to_id FROM user_relations WHERE to_id=?1")
        {
            Ok(s) => s,
            Err(_) => return edges,
        };
        let rows = stmt.query_map(params![entity_id], |row| {
            Ok(GraphEdge {
                from_id: row.get(0)?,
                relation_type: row.get(1)?,
                to_id: row.get(2)?,
            })
        });
        if let Ok(rows) = rows {
            for r in rows.flatten() {
                edges.push(r);
            }
        }

        edges
    }

    fn search_user_entities(&self, query: &str, limit: usize) -> Vec<UserEntity> {
        let conn = match self.conn.lock() {
            Ok(c) => c,
            Err(_) => return Vec::new(),
        };
        let mut stmt = match conn.prepare(
            "SELECT u.id, u.title, u.content, u.author, u.confidence, u.evidence_count,
                    u.last_validated, u.tags, u.relations, u.created_at, u.updated_at, u.link_provenance
             FROM user_entities_fts f
             JOIN user_entities u ON u.rowid = f.rowid
             WHERE user_entities_fts MATCH ?1
             ORDER BY rank
             LIMIT ?2",
        ) {
            Ok(s) => s,
            Err(_) => return Vec::new(),
        };
        let rows = stmt.query_map(params![query, limit as i64], row_to_user_entity);
        match rows {
            Ok(r) => r.filter_map(|x| x.ok()).collect(),
            Err(_) => Vec::new(),
        }
    }

    fn compute_correlations(&self, insight_id: &str) -> Vec<CorrelationScore> {
        let conn = match self.conn.lock() {
            Ok(c) => c,
            Err(_) => return Vec::new(),
        };
        // Get the target's neighbor targets as a flat set
        let target_targets: std::collections::HashSet<String> = {
            let mut stmt = match conn.prepare("SELECT to_id FROM user_relations WHERE from_id=?1") {
                Ok(s) => s,
                Err(_) => return Vec::new(),
            };
            let rows = stmt.query_map(params![insight_id], |row| row.get::<_, String>(0));
            match rows {
                Ok(r) => r.filter_map(|v| v.ok()).collect(),
                Err(_) => return Vec::new(),
            }
        };

        // Single query: get all other user entities with their neighbor targets
        // using a self-JOIN to compute Jaccard in SQL-adjacent logic
        let mut stmt = match conn.prepare(
            "SELECT r1.from_id, r1.to_id
             FROM user_relations r1
             WHERE r1.from_id != ?1",
        ) {
            Ok(s) => s,
            Err(_) => return Vec::new(),
        };
        let rows = stmt.query_map(params![insight_id], |row| {
            let from: String = row.get(0)?;
            let to: String = row.get(1)?;
            Ok((from, to))
        });

        // Group by entity: entity_id -> set of targets
        let mut entity_targets: HashMap<String, std::collections::HashSet<String>> = HashMap::new();
        if let Ok(rows) = rows {
            for r in rows.flatten() {
                entity_targets.entry(r.0).or_default().insert(r.1);
            }
        }

        let mut scores = Vec::new();
        for (other_id, other_targets) in &entity_targets {
            let intersection = target_targets.intersection(other_targets).count();
            let union = target_targets.union(other_targets).count();
            let graph_prox = if union == 0 {
                0.0
            } else {
                intersection as f64 / union as f64
            };

            let combined = 0.4 * graph_prox + 0.2 * 0.5;
            if combined > 0.05 {
                scores.push(CorrelationScore {
                    insight_id: other_id.clone(),
                    semantic: 0.0,
                    graph_proximity: graph_prox,
                    temporal: 0.5,
                    combined,
                });
            }
        }

        scores.sort_by(|a, b| {
            b.combined
                .partial_cmp(&a.combined)
                .unwrap_or(std::cmp::Ordering::Equal)
        });
        scores
    }

    fn store_embedding(&self, entity_id: &str, embedding: &[f32]) -> Result<(), String> {
        let conn = self
            .conn
            .lock()
            .map_err(|e| format!("mutex poisoned: {e}"))?;
        let bytes: Vec<u8> = embedding.iter().flat_map(|f| f.to_le_bytes()).collect();
        conn.execute(
            "INSERT OR REPLACE INTO user_embeddings (entity_id, embedding) VALUES (?1, ?2)",
            params![entity_id, bytes],
        )
        .map_err(|e| format!("store embedding: {e}"))?;
        Ok(())
    }

    fn get_embedding(&self, entity_id: &str) -> Option<Vec<f32>> {
        let conn = self.conn.lock().ok()?;
        let blob: Vec<u8> = conn
            .query_row(
                "SELECT embedding FROM user_embeddings WHERE entity_id=?1",
                params![entity_id],
                |row| row.get(0),
            )
            .ok()?;
        blob.chunks_exact(4)
            .map(|chunk| {
                let bytes: [u8; 4] = [chunk[0], chunk[1], chunk[2], chunk[3]];
                f32::from_le_bytes(bytes)
            })
            .collect::<Vec<_>>()
            .into()
    }

    fn user_entity_count(&self) -> usize {
        let conn = match self.conn.lock() {
            Ok(c) => c,
            Err(_) => return 0,
        };
        conn.query_row("SELECT COUNT(*) FROM user_entities", [], |row| {
            row.get::<_, i64>(0)
        })
        .ok()
        .and_then(|c| usize::try_from(c).ok())
        .unwrap_or(0)
    }

    /// Override: use atomic SQLite sequence counter instead of scanning.
    fn next_insight_id(&self) -> Result<String, String> {
        self.next_insight_id_atomic()
    }
}

// ---------------------------------------------------------------------------
// Tests
// ---------------------------------------------------------------------------

#[cfg(test)]
mod tests {
    use super::*;

    fn make_entity(id: &str, title: &str) -> UserEntity {
        UserEntity {
            id: id.to_owned(),
            title: title.to_owned(),
            content: format!("Content for {title}"),
            author: "test".to_owned(),
            confidence: 0.5,
            evidence_count: 0,
            last_validated: String::new(),
            tags: vec!["test".to_owned()],
            relations: HashMap::new(),
            link_provenance: HashMap::new(),
            created_at: "2026-01-01T00:00:00Z".to_owned(),
            updated_at: "2026-01-01T00:00:00Z".to_owned(),
        }
    }

    #[test]
    fn open_in_memory_initializes_schema() {
        let store = UserGraphStore::open_in_memory().unwrap();
        assert_eq!(store.user_entity_count(), 0);
    }

    #[test]
    fn add_and_get_entity() {
        let store = UserGraphStore::open_in_memory().unwrap();
        let entity = make_entity("TK-001", "Test Insight");
        store.add_entity(entity).unwrap();
        assert_eq!(store.user_entity_count(), 1);

        let retrieved = store.get_user_entity("TK-001").unwrap();
        assert_eq!(retrieved.title, "Test Insight");
        assert_eq!(retrieved.author, "test");
    }

    #[test]
    fn add_duplicate_entity_fails() {
        let store = UserGraphStore::open_in_memory().unwrap();
        store.add_entity(make_entity("TK-001", "First")).unwrap();
        let result = store.add_entity(make_entity("TK-001", "Second"));
        assert!(result.is_err());
    }

    #[test]
    fn update_entity() {
        let store = UserGraphStore::open_in_memory().unwrap();
        store.add_entity(make_entity("TK-001", "Original")).unwrap();
        let mut updated = make_entity("TK-001", "Updated");
        updated.confidence = 0.9;
        store.update_entity("TK-001", updated).unwrap();

        let retrieved = store.get_user_entity("TK-001").unwrap();
        assert_eq!(retrieved.title, "Updated");
        assert!((retrieved.confidence - 0.9).abs() < f64::EPSILON);
    }

    #[test]
    fn remove_entity_cascades() {
        let store = UserGraphStore::open_in_memory().unwrap();
        store.add_entity(make_entity("TK-001", "A")).unwrap();
        store.add_entity(make_entity("TK-002", "B")).unwrap();
        store
            .add_relation("TK-001", "derives_from", "TK-002")
            .unwrap();

        store.remove_entity("TK-001").unwrap();
        assert_eq!(store.user_entity_count(), 1);
        assert!(store.get_user_entity("TK-001").is_none());
        // TK-002 still exists but relation to TK-001 is gone
        let neighbors = store.get_user_neighbors("TK-002", None);
        assert!(neighbors.is_empty());
    }

    #[test]
    fn add_and_remove_relation() {
        let store = UserGraphStore::open_in_memory().unwrap();
        store.add_entity(make_entity("TK-001", "A")).unwrap();
        store.add_entity(make_entity("TK-002", "B")).unwrap();
        store
            .add_relation("TK-001", "derives_from", "TK-002")
            .unwrap();

        let neighbors = store.get_user_neighbors("TK-001", Some("derives_from"));
        assert_eq!(neighbors, vec!["TK-002"]);

        store
            .remove_relation("TK-001", "derives_from", "TK-002")
            .unwrap();
        let neighbors = store.get_user_neighbors("TK-001", Some("derives_from"));
        assert!(neighbors.is_empty());
    }

    #[test]
    fn search_by_keyword() {
        let store = UserGraphStore::open_in_memory().unwrap();
        let mut entity = make_entity("TK-001", "Strategy Pattern Decision");
        entity.content = "We decided to use Strategy for payment processing".to_owned();
        entity.tags = vec!["decision".to_owned(), "payment".to_owned()];
        store.add_entity(entity).unwrap();

        let results = store.search_user_entities("Strategy payment", 10);
        assert_eq!(results.len(), 1);
        assert_eq!(results[0].id, "TK-001");
    }

    #[test]
    fn store_and_retrieve_embedding() {
        let store = UserGraphStore::open_in_memory().unwrap();
        store.add_entity(make_entity("TK-001", "Embedded")).unwrap();
        let embedding = vec![0.1, 0.2, 0.3, 0.4];
        store.store_embedding("TK-001", &embedding).unwrap();

        let retrieved = store.get_embedding("TK-001").unwrap();
        assert_eq!(retrieved.len(), 4);
        assert!((retrieved[0] - 0.1).abs() < f32::EPSILON);
    }

    #[test]
    fn compute_correlations_shared_neighbors() {
        let store = UserGraphStore::open_in_memory().unwrap();
        store.add_entity(make_entity("TK-001", "A")).unwrap();
        store.add_entity(make_entity("TK-002", "B")).unwrap();
        // Both derive from same canonical entity
        store
            .add_relation("TK-001", "derives_from", "DP-005")
            .unwrap();
        store
            .add_relation("TK-002", "derives_from", "DP-005")
            .unwrap();

        let correlations = store.compute_correlations("TK-001");
        assert_eq!(correlations.len(), 1);
        assert_eq!(correlations[0].insight_id, "TK-002");
        assert!(correlations[0].graph_proximity > 0.9);
    }

    #[test]
    fn get_all_edges() {
        let store = UserGraphStore::open_in_memory().unwrap();
        store.add_entity(make_entity("TK-001", "A")).unwrap();
        store.add_entity(make_entity("TK-002", "B")).unwrap();
        store
            .add_relation("TK-001", "derives_from", "DP-005")
            .unwrap();
        store
            .add_relation("TK-001", "relates_to", "TK-002")
            .unwrap();

        let edges = store.get_user_all_edges("TK-001");
        assert_eq!(edges.len(), 2);
    }
}