mnueron 0.3.0 → 0.4.0

package/dist/store/consolidator.js

@@ -0,0 +1,168 @@
+// ─────────────────────────────────────────────────────────────────────────────
+// P5 — Self-revising memory loop (phase 5a: detection only).
+//
+// Background scan that finds likely-duplicate memories and surfaces them as
+// proposals the user can act on. NO automatic writes — phase 5a is purely
+// observational. Phase 5b (LLM-proposed merges) and 5c (auto-merge on
+// high-confidence cases) layer on top later.
+//
+// Detection method: embedding cosine similarity between every pair of
+// memories within a sliding window. We don't N² across the whole store —
+// that would melt. Instead:
+//
+//   1. Walk memories sorted by created_at DESC.
+//   2. For each memory M, vector-search the top-K nearest neighbors.
+//   3. Filter to candidates with cosine >= DUPLICATE_THRESHOLD.
+//   4. Emit one `ConsolidationProposal` per (M, neighbor) pair the
+//      reviewer hasn't already seen or actioned.
+//
+// State persists in `consolidation_proposals`:
+//   - status: 'pending' | 'approved' | 'rejected'
+//   - kind:   'duplicate' (5a) | 'contradiction' (5b) | 'stale' (5b)
+//   - score:  similarity that triggered the proposal
+// Action history per proposal is its own table so we can audit who/when
+// in a multi-user (hosted) context. The local store keeps it simple.
+//
+// Safety: phase 5a never deletes or mutates memories. The only mutation
+// is INSERTs into consolidation_proposals.
+// ─────────────────────────────────────────────────────────────────────────────
+import { randomUUID } from 'node:crypto';
+/** Cosine threshold for emitting a duplicate proposal. Tuned conservative —
+ *  we'd rather miss some duplicates than swamp the user with false positives. */
+const DUPLICATE_THRESHOLD = 0.92;
+/** Top-K neighbors examined per memory. Larger = catches more duplicates,
+ *  smaller = faster scan. K=5 covers ~all real-world duplicates per sampling. */
+const NEIGHBORS_PER_MEMORY = 5;
+/** Default cap on how many memories to scan per invocation. */
+const DEFAULT_SCAN_LIMIT = 200;
+/** Schema bootstrap. Idempotent. Called from local.ts migrate(). */
+export function ensureConsolidationSchema(db) {
+    db.exec(`
+    CREATE TABLE IF NOT EXISTS consolidation_proposals (
+      id            TEXT PRIMARY KEY,
+      kind          TEXT NOT NULL,
+      memory_a_id   TEXT NOT NULL,
+      memory_b_id   TEXT NOT NULL,
+      score         REAL NOT NULL,
+      status        TEXT NOT NULL DEFAULT 'pending',
+      note          TEXT,
+      proposed_at   INTEGER NOT NULL,
+      reviewed_at   INTEGER
+    );
+    CREATE INDEX IF NOT EXISTS idx_proposals_status
+      ON consolidation_proposals(status, proposed_at DESC);
+    CREATE INDEX IF NOT EXISTS idx_proposals_memory_a
+      ON consolidation_proposals(memory_a_id);
+    CREATE INDEX IF NOT EXISTS idx_proposals_memory_b
+      ON consolidation_proposals(memory_b_id);
+    -- Stop a duplicate proposal from being re-created on every scan.
+    CREATE UNIQUE INDEX IF NOT EXISTS uq_proposals_pair_kind
+      ON consolidation_proposals(memory_a_id, memory_b_id, kind);
+  `);
+}
+/**
+ * Phase 5a — pure detection pass. Walks memories in reverse chronological
+ * order, vector-searches top-K neighbors per memory, and inserts
+ * 'duplicate' proposals for any pair with similarity ≥ threshold.
+ *
+ * Idempotent via the (memory_a_id, memory_b_id, kind) unique index — if a
+ * proposal already exists it's silently retained (counted as
+ * `proposalsAlreadyKnown`).
+ *
+ * Returns counts so the CLI can print a one-line summary.
+ */
+export async function detectDuplicates(db, vecAvailable, opts = {}) {
+    if (!vecAvailable) {
+        // Without vectors we have no similarity signal — bail.
+        return { scanned: 0, proposalsCreated: 0, proposalsAlreadyKnown: 0 };
+    }
+    const limit = Math.max(1, Math.min(opts.limit ?? DEFAULT_SCAN_LIMIT, 5000));
+    const threshold = clamp01(opts.threshold ?? DUPLICATE_THRESHOLD);
+    // Pull seed memories. We need their stored embedding bytes from the vec
+    // table so we can use them as the search probe.
+    const seedRows = (opts.namespace
+        ? db.prepare(`SELECT m.id, m.namespace
+             FROM memories m
+            WHERE m.namespace = ?
+            ORDER BY m.created_at DESC
+            LIMIT ?`).all(opts.namespace, limit)
+        : db.prepare(`SELECT m.id, m.namespace
+             FROM memories m
+            ORDER BY m.created_at DESC
+            LIMIT ?`).all(limit));
+    const insert = db.prepare(`INSERT INTO consolidation_proposals
+       (id, kind, memory_a_id, memory_b_id, score, status, note, proposed_at)
+     VALUES (?, ?, ?, ?, ?, ?, ?, ?)
+     ON CONFLICT(memory_a_id, memory_b_id, kind) DO NOTHING`);
+    const now = Date.now();
+    let scanned = 0;
+    let proposalsCreated = 0;
+    let proposalsAlreadyKnown = 0;
+    for (const seed of seedRows) {
+        scanned += 1;
+        // Read this seed's embedding from memories_vec.
+        const row = db
+            .prepare(`SELECT embedding FROM memories_vec WHERE memory_id = ?`)
+            .get(seed.id);
+        if (!row?.embedding)
+            continue;
+        // Top-K nearest neighbors (excluding the seed itself).
+        const neighbors = db
+            .prepare(`SELECT memory_id, distance
+           FROM memories_vec
+          WHERE embedding MATCH ?
+            AND k = ?
+          ORDER BY distance ASC`)
+            .all(row.embedding, NEIGHBORS_PER_MEMORY + 1);
+        for (const n of neighbors) {
+            if (n.memory_id === seed.id)
+                continue; // skip self
+            const sim = clamp01(1 - (n.distance * n.distance) / 2);
+            if (sim < threshold)
+                continue;
+            // Canonicalize the pair: lower id first, so (A, B) and (B, A) collapse.
+            const [a, b] = seed.id < n.memory_id ? [seed.id, n.memory_id] : [n.memory_id, seed.id];
+            const before = db
+                .prepare(`SELECT 1 FROM consolidation_proposals
+            WHERE memory_a_id = ? AND memory_b_id = ? AND kind = 'duplicate' LIMIT 1`)
+                .get(a, b);
+            if (before) {
+                proposalsAlreadyKnown += 1;
+                continue;
+            }
+            insert.run(randomUUID(), 'duplicate', a, b, sim, 'pending', null, now);
+            proposalsCreated += 1;
+        }
+    }
+    return { scanned, proposalsCreated, proposalsAlreadyKnown };
+}
+export function listProposals(db, opts = {}) {
+    const parts = ['1=1'];
+    const params = [];
+    if (opts.status) {
+        parts.push('status = ?');
+        params.push(opts.status);
+    }
+    if (opts.kind) {
+        parts.push('kind   = ?');
+        params.push(opts.kind);
+    }
+    const limit = Math.max(1, Math.min(opts.limit ?? 100, 1000));
+    const offset = Math.max(0, opts.offset ?? 0);
+    return db
+        .prepare(`SELECT * FROM consolidation_proposals
+        WHERE ${parts.join(' AND ')}
+        ORDER BY proposed_at DESC
+        LIMIT ? OFFSET ?`)
+        .all(...params, limit, offset);
+}
+export function reviewProposal(db, id, decision) {
+    db.prepare(`UPDATE consolidation_proposals SET status = ?, reviewed_at = ? WHERE id = ?`).run(decision, Date.now(), id);
+    return db
+        .prepare(`SELECT * FROM consolidation_proposals WHERE id = ?`)
+        .get(id) ?? null;
+}
+function clamp01(n) {
+    return n < 0 ? 0 : n > 1 ? 1 : n;
+}
+//# sourceMappingURL=consolidator.js.map

package/dist/store/consolidator.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"consolidator.js","sourceRoot":"","sources":["../../src/store/consolidator.ts"],"names":[],"mappings":"AAAA,gFAAgF;AAChF,6DAA6D;AAC7D,EAAE;AACF,4EAA4E;AAC5E,0EAA0E;AAC1E,sEAAsE;AACtE,6CAA6C;AAC7C,EAAE;AACF,sEAAsE;AACtE,yEAAyE;AACzE,4BAA4B;AAC5B,EAAE;AACF,gDAAgD;AAChD,qEAAqE;AACrE,gEAAgE;AAChE,mEAAmE;AACnE,iDAAiD;AACjD,EAAE;AACF,+CAA+C;AAC/C,kDAAkD;AAClD,qEAAqE;AACrE,qDAAqD;AACrD,wEAAwE;AACxE,qEAAqE;AACrE,EAAE;AACF,wEAAwE;AACxE,2CAA2C;AAC3C,gFAAgF;AAGhF,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAEzC;iFACiF;AACjF,MAAM,mBAAmB,GAAG,IAAI,CAAC;AACjC;iFACiF;AACjF,MAAM,oBAAoB,GAAG,CAAC,CAAC;AAC/B,+DAA+D;AAC/D,MAAM,kBAAkB,GAAG,GAAG,CAAC;AAmB/B,oEAAoE;AACpE,MAAM,UAAU,yBAAyB,CAAC,EAAqB;IAC7D,EAAE,CAAC,IAAI,CAAC;;;;;;;;;;;;;;;;;;;;;GAqBP,CAAC,CAAC;AACL,CAAC;AAiBD;;;;;;;;;;GAUG;AACH,MAAM,CAAC,KAAK,UAAU,gBAAgB,CACpC,EAAqB,EACrB,YAAqB,EACrB,OAAoB,EAAE;IAEtB,IAAI,CAAC,YAAY,EAAE,CAAC;QAClB,uDAAuD;QACvD,OAAO,EAAE,OAAO,EAAE,CAAC,EAAE,gBAAgB,EAAE,CAAC,EAAE,qBAAqB,EAAE,CAAC,EAAE,CAAC;IACvE,CAAC;IACD,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,KAAK,IAAI,kBAAkB,EAAE,IAAI,CAAC,CAAC,CAAC;IAC5E,MAAM,SAAS,GAAG,OAAO,CAAC,IAAI,CAAC,SAAS,IAAI,mBAAmB,CAAC,CAAC;IAEjE,wEAAwE;IACxE,gDAAgD;IAChD,MAAM,QAAQ,GAAG,CACf,IAAI,CAAC,SAAS;QACZ,CAAC,CAAC,EAAE,CAAC,OAAO,CACR;;;;oBAIU,CACX,CAAC,GAAG,CAAC,IAAI,CAAC,SAAS,EAAE,KAAK,CAAC;QAC9B,CAAC,CAAC,EAAE,CAAC,OAAO,CACR;;;oBAGU,CACX,CAAC,GAAG,CAAC,KAAK,CAAC,CAC2B,CAAC;IAE9C,MAAM,MAAM,GAAG,EAAE,CAAC,OAAO,CAGvB;;;4DAGwD,CACzD,CAAC;IACF,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;IAEvB,IAAI,OAAO,GAAG,CAAC,CAAC;IAChB,IAAI,gBAAgB,GAAG,CAAC,CAAC;IACzB,IAAI,qBAAqB,GAAG,CAAC,CAAC;IAE9B,KAAK,MAAM,IAAI,IAAI,QAAQ,EAAE,CAAC;QAC5B,OAAO,IAAI,CAAC,CAAC;QAEb,gDAAgD;QAChD,MAAM,GAAG,GAAG,EAAE;aACX,OAAO,CAAC,wDAAwD,CAAC;aACjE,GAAG,CAAC,IAAI,CAAC,EAAE,CAAuC,CAAC;QACtD,IAAI,CAAC,GAAG,EAAE,SAAS;YAAE,SAAS;QAE9B,uDAAuD;QACvD,MAAM,SAAS,GAAG,EAAE;aACjB,OAAO,CACN;;;;gCAIwB,CACzB;aACA,GAAG,CAAC,GAAG,CAAC,SAAS,EAAE,oBAAoB,GAAG,CAAC,CAG1C,CAAC;QAEL,KAAK,MAAM,CAAC,IAAI,SAAS,EAAE,CAAC;YAC1B,IAAI,CAAC,CAAC,SAAS,KAAK,IAAI,CAAC,EAAE;gBAAE,SAAS,CAAC,YAAY;YACnD,MAAM,GAAG,GAAG,OAAO,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC;YACvD,IAAI,GAAG,GAAG,SAAS;gBAAE,SAAS;YAE9B,wEAAwE;YACxE,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,GAAG,IAAI,CAAC,EAAE,GAAG,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,EAAE,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,EAAE,IAAI,CAAC,EAAE,CAAC,CAAC;YAEvF,MAAM,MAAM,GAAG,EAAE;iBACd,OAAO,CACN;qFAC2E,CAC5E;iBACA,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;YACb,IAAI,MAAM,EAAE,CAAC;gBACX,qBAAqB,IAAI,CAAC,CAAC;gBAC3B,SAAS;YACX,CAAC;YACD,MAAM,CAAC,GAAG,CACR,UAAU,EAAE,EACZ,WAAW,EACX,CAAC,EACD,CAAC,EACD,GAAG,EACH,SAAS,EACT,IAAI,EACJ,GAAG,CACJ,CAAC;YACF,gBAAgB,IAAI,CAAC,CAAC;QACxB,CAAC;IACH,CAAC;IAED,OAAO,EAAE,OAAO,EAAE,gBAAgB,EAAE,qBAAqB,EAAE,CAAC;AAC9D,CAAC;AASD,MAAM,UAAU,aAAa,CAC3B,EAAqB,EACrB,OAA4B,EAAE;IAE9B,MAAM,KAAK,GAAa,CAAC,KAAK,CAAC,CAAC;IAChC,MAAM,MAAM,GAAc,EAAE,CAAC;IAC7B,IAAI,IAAI,CAAC,MAAM,EAAE,CAAC;QAAC,KAAK,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC;QAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IAAC,CAAC;IACxE,IAAI,IAAI,CAAC,IAAI,EAAI,CAAC;QAAC,KAAK,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC;QAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAAG,CAAC;IACxE,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,KAAK,IAAI,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC;IAC7D,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,MAAM,IAAI,CAAC,CAAC,CAAC;IAC7C,OAAO,EAAE;SACN,OAAO,CACN;gBACU,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC;;yBAEV,CACpB;SACA,GAAG,CAAC,GAAG,MAAM,EAAE,KAAK,EAAE,MAAM,CAA4B,CAAC;AAC9D,CAAC;AAED,MAAM,UAAU,cAAc,CAC5B,EAAqB,EACrB,EAAU,EACV,QAAiC;IAEjC,EAAE,CAAC,OAAO,CACR,6EAA6E,CAC9E,CAAC,GAAG,CAAC,QAAQ,EAAE,IAAI,CAAC,GAAG,EAAE,EAAE,EAAE,CAAC,CAAC;IAChC,OAAQ,EAAE;SACP,OAAO,CAAC,oDAAoD,CAAC;SAC7D,GAAG,CAAC,EAAE,CAAuC,IAAI,IAAI,CAAC;AAC3D,CAAC;AAED,SAAS,OAAO,CAAC,CAAS;IACxB,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AACnC,CAAC"}

package/dist/store/entity-extractor.js

@@ -0,0 +1,283 @@
+/**
+ * P1 — Entity extraction for the local CLI / SQLite provider.
+ *
+ * Mirrors the hosted-backend entity-extractor (see ai-boilerplate-pro/src/lib/
+ * entity-extractor.ts) but uses raw fetch for both providers to avoid pulling
+ * the Anthropic SDK into the CLI's dependency tree. The CLI ships to npm and
+ * gets installed globally by users, so we keep deps lean.
+ *
+ * Provider precedence (per-call BYOK keys never persisted):
+ *   1. metadata.byok_anthropic_key  -> Claude Haiku via raw fetch
+ *   2. metadata.byok_openai_key     -> gpt-4o-mini via raw fetch
+ *   3. ANTHROPIC_API_KEY env var    -> Claude Haiku
+ *   4. OPENAI_API_KEY env var       -> gpt-4o-mini
+ *   5. none -> return [], save proceeds with no entities (fail-open)
+ *
+ * Gating: opt-in. set `metadata.extract_entities: true` per-call, or
+ * the `MNUERON_ENABLE_ENTITY_EXTRACTION` env var globally.
+ *
+ * Fail-open across the board. Save never blocks on extraction.
+ */
+const MIN_LENGTH_CHARS = 200;
+const ANTHROPIC_MODEL = 'claude-haiku-4-5';
+const OPENAI_MODEL = 'gpt-4o-mini';
+const MAX_CONTENT_CHARS = 12000;
+const MAX_ENTITIES = 25;
+const TIMEOUT_MS = 30000;
+export const ENTITY_EXTRACTION_ENABLED = (process.env.MNUERON_ENABLE_ENTITY_EXTRACTION ?? '').toLowerCase() === 'true';
+/**
+ * Per-call gate. Same shape as hosted-side `shouldExtractEntities`.
+ *
+ * Explicit per-call opt-in (metadata.extract_entities: true OR BYOK key)
+ * always runs, even on short content. The length floor only applies to
+ * the env-var default path, to keep that from burning money on noise.
+ */
+export function shouldExtractEntities(contentLen, metadata, minChars = MIN_LENGTH_CHARS) {
+    if (metadata?.extract_entities === true)
+        return true;
+    const a = metadata?.byok_anthropic_key;
+    if (typeof a === 'string' && a.length > 0)
+        return true;
+    const o = metadata?.byok_openai_key;
+    if (typeof o === 'string' && o.length > 0)
+        return true;
+    if (ENTITY_EXTRACTION_ENABLED && contentLen >= minChars)
+        return true;
+    return false;
+}
+/**
+ * Pull a structured entity list from `content`. Tries each provider in
+ * order until one returns results. Returns `[]` on every failure.
+ */
+export async function extractEntities(content, opts = {}) {
+    const min = opts.minChars ?? MIN_LENGTH_CHARS;
+    if (!content || content.length < min)
+        return [];
+    const trimmed = content.slice(0, MAX_CONTENT_CHARS);
+    try {
+        if (opts.anthropicKey) {
+            const out = await extractViaAnthropic(trimmed, opts.anthropicKey);
+            if (out.length > 0)
+                return cap(out);
+        }
+        if (opts.openaiKey) {
+            const out = await extractViaOpenAI(trimmed, opts.openaiKey);
+            if (out.length > 0)
+                return cap(out);
+        }
+        if (process.env.ANTHROPIC_API_KEY) {
+            const out = await extractViaAnthropic(trimmed, process.env.ANTHROPIC_API_KEY);
+            if (out.length > 0)
+                return cap(out);
+        }
+        if (process.env.OPENAI_API_KEY) {
+            const out = await extractViaOpenAI(trimmed, process.env.OPENAI_API_KEY);
+            if (out.length > 0)
+                return cap(out);
+        }
+    }
+    catch (e) {
+        console.warn('[mnueron/entity-extractor]', e instanceof Error ? e.message : e);
+    }
+    return [];
+}
+const SYSTEM_PROMPT = [
+    'You extract structured entities from memory text that will be stored',
+    "alongside an AI agent's long-term memory. The goal is a list of",
+    'recurring, identifiable, *referenceable* things — not a list of every',
+    'noun in the text.',
+    '',
+    'For each entity found, output:',
+    '',
+    '  - name: the canonical display form (proper noun preferred)',
+    '  - type: one of "person", "organization", "project", "technology",',
+    '          "place", "decision", "event", "concept", "other"',
+    '  - context: a short phrase (<= 80 chars) from the source that disambiguates',
+    '             this entity from others with the same name. Optional.',
+    '',
+    'STRONG inclusion criteria — only emit an entity if it would survive these:',
+    '  - Is it a proper noun, a named system, or a clearly bounded concept?',
+    '    (YES: "Stripe", "Q3 roadmap", "PostgreSQL", "deprecate v1 API decision")',
+    '    (NO:  "Step 1 view file", "DB-backed settings", "server-side validation")',
+    '  - Will another memory plausibly reference this same thing again?',
+    '    If you can\'t imagine a second memory mentioning it, skip it.',
+    '  - Is it identifiable on its own? "the API" is not — "Stripe API v2" is.',
+    '',
+    'EXPLICIT EXCLUSIONS — never emit any of these:',
+    '  - Action phrases / implementation steps ("Step 1", "set up X", "fix Y",',
+    '    "DB-backed", "server-side validation", "role-based auth").',
+    '  - Generic technical terms used as common nouns ("the database",',
+    '    "the wizard", "the controller", "the API call").',
+    '  - File paths or hostnames as "place" — those are "technology".',
+    '  - Verbs or verb phrases as concepts.',
+    '  - Single-letter or all-lowercase one-word concepts ("PROD" can stay if',
+    '    it\'s clearly an environment label, but "qa" or "dev" alone — skip).',
+    '  - Anything mentioned only once in passing with no further context.',
+    '',
+    'Type guidance:',
+    '  - "place"      ONLY for real geographic locations (cities, offices,',
+    '                 countries). Hostnames, URLs, file paths → "technology".',
+    '  - "decision"   for explicit choices made ("we decided to use X").',
+    '                 NOT for implementation steps or recommendations in passing.',
+    '  - "project"    for named multi-task efforts ("Q3 roadmap", "Auth Rewrite").',
+    '                 NOT for one-off files or single UI screens.',
+    '  - "concept"    sparingly — only for abstract ideas referenced multiple times.',
+    '',
+    'Cap output at 25 entities; if more exist, pick the most important.',
+    'If fewer than 2 high-quality entities exist, return [] — better to return',
+    'nothing than to fill the list with noise.',
+    '',
+    'Respond with STRICT JSON. No prose, no preamble, no markdown.',
+    'Schema:',
+    '  { "entities": [ { "name": "...", "type": "...", "context": "..." }, ... ] }',
+].join('\n');
+async function extractViaAnthropic(content, apiKey) {
+    const ctrl = new AbortController();
+    const timer = setTimeout(() => ctrl.abort(), TIMEOUT_MS);
+    try {
+        const resp = await fetch('https://api.anthropic.com/v1/messages', {
+            method: 'POST',
+            headers: {
+                'x-api-key': apiKey,
+                'anthropic-version': '2023-06-01',
+                'Content-Type': 'application/json',
+            },
+            body: JSON.stringify({
+                model: ANTHROPIC_MODEL,
+                max_tokens: 1500,
+                temperature: 0.1,
+                system: SYSTEM_PROMPT,
+                messages: [{ role: 'user', content }],
+            }),
+            signal: ctrl.signal,
+        });
+        if (!resp.ok) {
+            const body = await resp.text();
+            console.warn('[mnueron/entity-extractor/anthropic] HTTP ' + resp.status + ': ' + body.slice(0, 200));
+            return [];
+        }
+        const data = (await resp.json());
+        const text = (data.content ?? [])
+            .filter((b) => b.type === 'text')
+            .map((b) => b.text ?? '')
+            .join('');
+        return parseEntities(text);
+    }
+    catch (e) {
+        console.warn('[mnueron/entity-extractor/anthropic]', e instanceof Error ? e.message : e);
+        return [];
+    }
+    finally {
+        clearTimeout(timer);
+    }
+}
+async function extractViaOpenAI(content, apiKey) {
+    const ctrl = new AbortController();
+    const timer = setTimeout(() => ctrl.abort(), TIMEOUT_MS);
+    try {
+        const resp = await fetch('https://api.openai.com/v1/chat/completions', {
+            method: 'POST',
+            headers: {
+                Authorization: 'Bearer ' + apiKey,
+                'Content-Type': 'application/json',
+            },
+            body: JSON.stringify({
+                model: OPENAI_MODEL,
+                max_tokens: 1500,
+                temperature: 0.1,
+                response_format: { type: 'json_object' },
+                messages: [
+                    { role: 'system', content: SYSTEM_PROMPT },
+                    { role: 'user', content },
+                ],
+            }),
+            signal: ctrl.signal,
+        });
+        if (!resp.ok) {
+            const body = await resp.text();
+            console.warn('[mnueron/entity-extractor/openai] HTTP ' + resp.status + ': ' + body.slice(0, 200));
+            return [];
+        }
+        const data = (await resp.json());
+        return parseEntities(data.choices?.[0]?.message?.content ?? '');
+    }
+    catch (e) {
+        console.warn('[mnueron/entity-extractor/openai]', e instanceof Error ? e.message : e);
+        return [];
+    }
+    finally {
+        clearTimeout(timer);
+    }
+}
+/**
+ * Resilient JSON parse:
+ *   - strips ```json ... ``` fences
+ *   - skips leading prose ("Here are the entities:\n{...}")
+ *   - accepts entities as bare array OR under `entities`/`results`/`items`
+ */
+function parseEntities(raw) {
+    if (!raw)
+        return [];
+    let s = raw.trim();
+    if (s.startsWith('```')) {
+        s = s.replace(/^```(?:json)?\s*/i, '').replace(/\s*```$/i, '');
+    }
+    const start = s.search(/[\[{]/);
+    if (start > 0)
+        s = s.slice(start);
+    let parsed;
+    try {
+        parsed = JSON.parse(s);
+    }
+    catch {
+        return [];
+    }
+    const arr = pickEntitiesArray(parsed);
+    if (!Array.isArray(arr))
+        return [];
+    const out = [];
+    for (const raw of arr) {
+        if (!raw || typeof raw !== 'object')
+            continue;
+        const r = raw;
+        const name = typeof r.name === 'string' ? r.name.trim() : '';
+        if (!name)
+            continue;
+        const type = typeof r.type === 'string' ? r.type.trim() : 'other';
+        const context = typeof r.context === 'string' ? r.context.trim().slice(0, 80) : undefined;
+        out.push({ name, type, context, canonical_id: null });
+    }
+    return dedupe(out);
+}
+function pickEntitiesArray(parsed) {
+    if (Array.isArray(parsed))
+        return parsed;
+    if (parsed && typeof parsed === 'object') {
+        const p = parsed;
+        if (Array.isArray(p.entities))
+            return p.entities;
+        if (Array.isArray(p.results))
+            return p.results;
+        if (Array.isArray(p.items))
+            return p.items;
+        if (Array.isArray(p.extracted))
+            return p.extracted;
+    }
+    return null;
+}
+function dedupe(items) {
+    const seen = new Set();
+    const out = [];
+    for (const item of items) {
+        const key = item.name.toLowerCase();
+        if (seen.has(key))
+            continue;
+        seen.add(key);
+        out.push(item);
+    }
+    return out;
+}
+function cap(items) {
+    return items.length > MAX_ENTITIES ? items.slice(0, MAX_ENTITIES) : items;
+}
+//# sourceMappingURL=entity-extractor.js.map

package/dist/store/entity-extractor.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"entity-extractor.js","sourceRoot":"","sources":["../../src/store/entity-extractor.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;GAmBG;AAEH,MAAM,gBAAgB,GAAG,GAAG,CAAC;AAC7B,MAAM,eAAe,GAAG,kBAAkB,CAAC;AAC3C,MAAM,YAAY,GAAG,aAAa,CAAC;AACnC,MAAM,iBAAiB,GAAG,KAAK,CAAC;AAChC,MAAM,YAAY,GAAG,EAAE,CAAC;AACxB,MAAM,UAAU,GAAG,KAAK,CAAC;AAezB,MAAM,CAAC,MAAM,yBAAyB,GACpC,CAAC,OAAO,CAAC,GAAG,CAAC,gCAAgC,IAAI,EAAE,CAAC,CAAC,WAAW,EAAE,KAAK,MAAM,CAAC;AAEhF;;;;;;GAMG;AACH,MAAM,UAAU,qBAAqB,CACnC,UAAkB,EAClB,QAA6C,EAC7C,QAAQ,GAAG,gBAAgB;IAE3B,IAAI,QAAQ,EAAE,gBAAgB,KAAK,IAAI;QAAE,OAAO,IAAI,CAAC;IACrD,MAAM,CAAC,GAAG,QAAQ,EAAE,kBAAkB,CAAC;IACvC,IAAI,OAAO,CAAC,KAAK,QAAQ,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,IAAI,CAAC;IACvD,MAAM,CAAC,GAAG,QAAQ,EAAE,eAAe,CAAC;IACpC,IAAI,OAAO,CAAC,KAAK,QAAQ,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,IAAI,CAAC;IACvD,IAAI,yBAAyB,IAAI,UAAU,IAAI,QAAQ;QAAE,OAAO,IAAI,CAAC;IACrE,OAAO,KAAK,CAAC;AACf,CAAC;AAED;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,eAAe,CACnC,OAAe,EACf,OAAuB,EAAE;IAEzB,MAAM,GAAG,GAAG,IAAI,CAAC,QAAQ,IAAI,gBAAgB,CAAC;IAC9C,IAAI,CAAC,OAAO,IAAI,OAAO,CAAC,MAAM,GAAG,GAAG;QAAE,OAAO,EAAE,CAAC;IAChD,MAAM,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,iBAAiB,CAAC,CAAC;IAEpD,IAAI,CAAC;QACH,IAAI,IAAI,CAAC,YAAY,EAAE,CAAC;YACtB,MAAM,GAAG,GAAG,MAAM,mBAAmB,CAAC,OAAO,EAAE,IAAI,CAAC,YAAY,CAAC,CAAC;YAClE,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC;gBAAE,OAAO,GAAG,CAAC,GAAG,CAAC,CAAC;QACtC,CAAC;QACD,IAAI,IAAI,CAAC,SAAS,EAAE,CAAC;YACnB,MAAM,GAAG,GAAG,MAAM,gBAAgB,CAAC,OAAO,EAAE,IAAI,CAAC,SAAS,CAAC,CAAC;YAC5D,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC;gBAAE,OAAO,GAAG,CAAC,GAAG,CAAC,CAAC;QACtC,CAAC;QACD,IAAI,OAAO,CAAC,GAAG,CAAC,iBAAiB,EAAE,CAAC;YAClC,MAAM,GAAG,GAAG,MAAM,mBAAmB,CAAC,OAAO,EAAE,OAAO,CAAC,GAAG,CAAC,iBAAiB,CAAC,CAAC;YAC9E,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC;gBAAE,OAAO,GAAG,CAAC,GAAG,CAAC,CAAC;QACtC,CAAC;QACD,IAAI,OAAO,CAAC,GAAG,CAAC,cAAc,EAAE,CAAC;YAC/B,MAAM,GAAG,GAAG,MAAM,gBAAgB,CAAC,OAAO,EAAE,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC,CAAC;YACxE,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC;gBAAE,OAAO,GAAG,CAAC,GAAG,CAAC,CAAC;QACtC,CAAC;IACH,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,OAAO,CAAC,IAAI,CAAC,4BAA4B,EAAE,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACjF,CAAC;IACD,OAAO,EAAE,CAAC;AACZ,CAAC;AAED,MAAM,aAAa,GAAG;IACpB,sEAAsE;IACtE,iEAAiE;IACjE,uEAAuE;IACvE,mBAAmB;IACnB,EAAE;IACF,gCAAgC;IAChC,EAAE;IACF,8DAA8D;IAC9D,qEAAqE;IACrE,4DAA4D;IAC5D,8EAA8E;IAC9E,oEAAoE;IACpE,EAAE;IACF,4EAA4E;IAC5E,wEAAwE;IACxE,8EAA8E;IAC9E,+EAA+E;IAC/E,oEAAoE;IACpE,mEAAmE;IACnE,2EAA2E;IAC3E,EAAE;IACF,gDAAgD;IAChD,2EAA2E;IAC3E,gEAAgE;IAChE,mEAAmE;IACnE,sDAAsD;IACtD,kEAAkE;IAClE,wCAAwC;IACxC,0EAA0E;IAC1E,0EAA0E;IAC1E,sEAAsE;IACtE,EAAE;IACF,gBAAgB;IAChB,uEAAuE;IACvE,0EAA0E;IAC1E,qEAAqE;IACrE,8EAA8E;IAC9E,+EAA+E;IAC/E,8DAA8D;IAC9D,iFAAiF;IACjF,EAAE;IACF,oEAAoE;IACpE,2EAA2E;IAC3E,2CAA2C;IAC3C,EAAE;IACF,+DAA+D;IAC/D,SAAS;IACT,+EAA+E;CAChF,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAEb,KAAK,UAAU,mBAAmB,CAChC,OAAe,EACf,MAAc;IAEd,MAAM,IAAI,GAAG,IAAI,eAAe,EAAE,CAAC;IACnC,MAAM,KAAK,GAAG,UAAU,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,KAAK,EAAE,EAAE,UAAU,CAAC,CAAC;IACzD,IAAI,CAAC;QACH,MAAM,IAAI,GAAG,MAAM,KAAK,CAAC,uCAAuC,EAAE;YAChE,MAAM,EAAE,MAAM;YACd,OAAO,EAAE;gBACP,WAAW,EAAE,MAAM;gBACnB,mBAAmB,EAAE,YAAY;gBACjC,cAAc,EAAE,kBAAkB;aACnC;YACD,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC;gBACnB,KAAK,EAAE,eAAe;gBACtB,UAAU,EAAE,IAAI;gBAChB,WAAW,EAAE,GAAG;gBAChB,MAAM,EAAE,aAAa;gBACrB,QAAQ,EAAE,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,CAAC;aACtC,CAAC;YACF,MAAM,EAAE,IAAI,CAAC,MAAM;SACpB,CAAC,CAAC;QACH,IAAI,CAAC,IAAI,CAAC,EAAE,EAAE,CAAC;YACb,MAAM,IAAI,GAAG,MAAM,IAAI,CAAC,IAAI,EAAE,CAAC;YAC/B,OAAO,CAAC,IAAI,CACV,4CAA4C,GAAG,IAAI,CAAC,MAAM,GAAG,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,CACvF,CAAC;YACF,OAAO,EAAE,CAAC;QACZ,CAAC;QACD,MAAM,IAAI,GAAG,CAAC,MAAM,IAAI,CAAC,IAAI,EAAE,CAE9B,CAAC;QACF,MAAM,IAAI,GAAG,CAAC,IAAI,CAAC,OAAO,IAAI,EAAE,CAAC;aAC9B,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,KAAK,MAAM,CAAC;aAChC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC;aACxB,IAAI,CAAC,EAAE,CAAC,CAAC;QACZ,OAAO,aAAa,CAAC,IAAI,CAAC,CAAC;IAC7B,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,OAAO,CAAC,IAAI,CAAC,sCAAsC,EAAE,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACzF,OAAO,EAAE,CAAC;IACZ,CAAC;YAAS,CAAC;QACT,YAAY,CAAC,KAAK,CAAC,CAAC;IACtB,CAAC;AACH,CAAC;AAED,KAAK,UAAU,gBAAgB,CAAC,OAAe,EAAE,MAAc;IAC7D,MAAM,IAAI,GAAG,IAAI,eAAe,EAAE,CAAC;IACnC,MAAM,KAAK,GAAG,UAAU,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,KAAK,EAAE,EAAE,UAAU,CAAC,CAAC;IACzD,IAAI,CAAC;QACH,MAAM,IAAI,GAAG,MAAM,KAAK,CAAC,4CAA4C,EAAE;YACrE,MAAM,EAAE,MAAM;YACd,OAAO,EAAE;gBACP,aAAa,EAAE,SAAS,GAAG,MAAM;gBACjC,cAAc,EAAE,kBAAkB;aACnC;YACD,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC;gBACnB,KAAK,EAAE,YAAY;gBACnB,UAAU,EAAE,IAAI;gBAChB,WAAW,EAAE,GAAG;gBAChB,eAAe,EAAE,EAAE,IAAI,EAAE,aAAa,EAAE;gBACxC,QAAQ,EAAE;oBACR,EAAE,IAAI,EAAE,QAAQ,EAAE,OAAO,EAAE,aAAa,EAAE;oBAC1C,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE;iBAC1B;aACF,CAAC;YACF,MAAM,EAAE,IAAI,CAAC,MAAM;SACpB,CAAC,CAAC;QACH,IAAI,CAAC,IAAI,CAAC,EAAE,EAAE,CAAC;YACb,MAAM,IAAI,GAAG,MAAM,IAAI,CAAC,IAAI,EAAE,CAAC;YAC/B,OAAO,CAAC,IAAI,CACV,yCAAyC,GAAG,IAAI,CAAC,MAAM,GAAG,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,CACpF,CAAC;YACF,OAAO,EAAE,CAAC;QACZ,CAAC;QACD,MAAM,IAAI,GAAG,CAAC,MAAM,IAAI,CAAC,IAAI,EAAE,CAE9B,CAAC;QACF,OAAO,aAAa,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,OAAO,IAAI,EAAE,CAAC,CAAC;IAClE,CAAC;IAAC,OAAO,CAAC,EAAE,CAAC;QACX,OAAO,CAAC,IAAI,CAAC,mCAAmC,EAAE,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACtF,OAAO,EAAE,CAAC;IACZ,CAAC;YAAS,CAAC;QACT,YAAY,CAAC,KAAK,CAAC,CAAC;IACtB,CAAC;AACH,CAAC;AAED;;;;;GAKG;AACH,SAAS,aAAa,CAAC,GAAW;IAChC,IAAI,CAAC,GAAG;QAAE,OAAO,EAAE,CAAC;IACpB,IAAI,CAAC,GAAG,GAAG,CAAC,IAAI,EAAE,CAAC;IACnB,IAAI,CAAC,CAAC,UAAU,CAAC,KAAK,CAAC,EAAE,CAAC;QACxB,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC,mBAAmB,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC;IACjE,CAAC;IACD,MAAM,KAAK,GAAG,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC;IAChC,IAAI,KAAK,GAAG,CAAC;QAAE,CAAC,GAAG,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;IAElC,IAAI,MAAe,CAAC;IACpB,IAAI,CAAC;QACH,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;IACzB,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,GAAG,GAAG,iBAAiB,CAAC,MAAM,CAAC,CAAC;IACtC,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC;QAAE,OAAO,EAAE,CAAC;IAEnC,MAAM,GAAG,GAAsB,EAAE,CAAC;IAClC,KAAK,MAAM,GAAG,IAAI,GAAG,EAAE,CAAC;QACtB,IAAI,CAAC,GAAG,IAAI,OAAO,GAAG,KAAK,QAAQ;YAAE,SAAS;QAC9C,MAAM,CAAC,GAAG,GAA8B,CAAC;QACzC,MAAM,IAAI,GAAG,OAAO,CAAC,CAAC,IAAI,KAAK,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC7D,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,MAAM,IAAI,GAAG,OAAO,CAAC,CAAC,IAAI,KAAK,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC;QAClE,MAAM,OAAO,GACX,OAAO,CAAC,CAAC,OAAO,KAAK,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;QAC5E,GAAG,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,OAAO,EAAE,YAAY,EAAE,IAAI,EAAE,CAAC,CAAC;IACxD,CAAC;IACD,OAAO,MAAM,CAAC,GAAG,CAAC,CAAC;AACrB,CAAC;AAED,SAAS,iBAAiB,CAAC,MAAe;IACxC,IAAI,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC;QAAE,OAAO,MAAM,CAAC;IACzC,IAAI,MAAM,IAAI,OAAO,MAAM,KAAK,QAAQ,EAAE,CAAC;QACzC,MAAM,CAAC,GAAG,MAAiC,CAAC;QAC5C,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,QAAQ,CAAC;YAAE,OAAO,CAAC,CAAC,QAAQ,CAAC;QACjD,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC;YAAE,OAAO,CAAC,CAAC,OAAO,CAAC;QAC/C,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC;YAAE,OAAO,CAAC,CAAC,KAAK,CAAC;QAC3C,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,SAAS,CAAC;YAAE,OAAO,CAAC,CAAC,SAAS,CAAC;IACrD,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,MAAM,CAAC,KAAwB;IACtC,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;IAC/B,MAAM,GAAG,GAAsB,EAAE,CAAC;IAClC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,GAAG,GAAG,IAAI,CAAC,IAAI,CAAC,WAAW,EAAE,CAAC;QACpC,IAAI,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;YAAE,SAAS;QAC5B,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QACd,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACjB,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAS,GAAG,CAAC,KAAwB;IACnC,OAAO,KAAK,CAAC,MAAM,GAAG,YAAY,CAAC,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,YAAY,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC;AAC5E,CAAC"}