akm-cli 0.9.17-alpha.4 → 0.9.17-alpha.6

package/dist/core/asset/asset-placement.js

@@ -26,7 +26,7 @@
 import fs from "node:fs";
 import path from "node:path";
 import { toPosix } from "../common.js";
-import { DERIVED_SUFFIX, SCRIPT_EXTENSIONS, WORKFLOW_EXTENSIONS } from "../recognition-util.js";
+import { SCRIPT_EXTENSIONS, WORKFLOW_EXTENSIONS } from "../recognition-util.js";
 const workflowSpec = {
     isRelevantFile: (fileName) => WORKFLOW_EXTENSIONS.includes(path.extname(fileName).toLowerCase()),
     toCanonicalName: (typeRoot, filePath) => {
@@ -245,21 +245,12 @@ export function assetPathForName(assetType, typeRoot, name) {
  * "default" alias is genuinely dual-owned: both `<dir>/.env` and
  * `<dir>/default.env` derive the same canonical name (`toCanonicalName`
  * above), so a physical-owner lookup must consider both without reading
- * either file. `memory` has a second, analogous duality (#882): a ref to
- * `<name>` may own either `<name>.md` or the LLM-inferred `<name>.derived.md`
- * twin — `.derived` is a provenance marker on the SAME identity, not part of
- * the name (see `resolveParentRef`/`isDerivedMemory` in
- * `commands/improve/memory/derived-ref.ts`, and the belief-edge identity
- * channel's own `memory:<name>.derived` refs). The plain `.md` file wins when
- * both exist, so it stays `primary` — first in the returned list — and every
- * caller here already prefers the first candidate that exists on disk. Every
- * other placement type has exactly one inverse spelling.
+ * either file. Every other placement type has exactly one inverse spelling —
+ * including `memory`: `<name>.derived.md` is a separate item that owns
+ * `<name>.derived`, never a second spelling of `<name>`.
  */
 export function assetPathCandidatesForName(assetType, typeRoot, name) {
     const primary = assetPathForName(assetType, typeRoot, name);
-    if (assetType === "memory" && !name.endsWith(DERIVED_SUFFIX)) {
-        return [primary, assetPathForName(assetType, typeRoot, `${name}${DERIVED_SUFFIX}`)];
-    }
     if (assetType !== "env")
         return [primary];
     const base = name === "default" ? "" : name.endsWith("/default") ? name.slice(0, -"default".length) : undefined;

package/dist/core/config/config.js

@@ -33,7 +33,7 @@ export { FEEDBACK_FAILURE_MODES } from "./config-schema.js";
  * combined prompt size well under common 8K/16K context windows (each body is
  * sliced to ~500 chars in the graph-extract prompt builder).
  */
-export const DEFAULT_GRAPH_EXTRACTION_BATCH_SIZE = 4;
+const DEFAULT_GRAPH_EXTRACTION_BATCH_SIZE = 4;
 /**
  * Approximate character budget per asset body inside a batched
  * graph-extraction prompt — used by {@link resolveBatchSize} to derive a

package/dist/core/config/schema/improve-processes.js

@@ -183,9 +183,10 @@ const MEMORY_INFERENCE_PROCESS_FIELDS = {
     cls: clsField,
 };
 /**
- * GraphExtraction process fields: improve-owned graph extraction scope and
- * batching. Passed to the invocation directly and never inherited from
- * standalone index.graph.
+ * GraphExtraction process fields: one strategy's graph extraction scope and
+ * batching. `includeTypes` and `batchSize` override `index.graph`'s
+ * `graphExtractionIncludeTypes` and `graphExtractionBatchSize`; unset, the
+ * pass reads those.
  */
 const GRAPH_EXTRACTION_PROCESS_FIELDS = {
     // #624 P2: when set, rank eligible files by utility_scores DESC and process

package/dist/core/config/schema/index-config.js

@@ -32,22 +32,11 @@ const INDEX_PASS_RETIRED_KEYS = new Set([
     "maxTokens",
     "capabilities",
 ]);
-const INDEX_PASS_KNOWN_KEYS = new Set([
-    "engine",
-    "model",
-    "timeoutMs",
-    "enabled",
-    "llm",
-    "graphExtractionBatchSize",
-    "graphExtractionIncludeTypes",
-    "lazyGraphExtraction",
-]);
 /**
- * Per-pass `index.<pass>` entry. Uses preprocess + manual validation so we can
- * emit targeted error messages ("Retired or misplaced engine setting",
- * "Unknown key `index.<pass>.<key>`")
- * instead of Zod's generic `Unrecognized key` / `Expected boolean, received
- * string` strings — keeps `akm` startup errors actionable.
+ * Per-pass `index.<pass>` entry. The preprocess names and drops the retired
+ * engine settings above with a targeted message. Any other unknown key is kept
+ * and named once by the config loader's schema walk, like an unknown key
+ * anywhere else in config.
  */
 export const IndexPassConfigSchema = z.preprocess((raw, ctx) => {
     if (typeof raw !== "object" || raw === null || Array.isArray(raw)) {
@@ -62,13 +51,6 @@ export const IndexPassConfigSchema = z.preprocess((raw, ctx) => {
             cleaned ??= { ...obj };
             delete cleaned[key];
         }
-        else if (!INDEX_PASS_KNOWN_KEYS.has(key)) {
-            warnOnce(`index-pass:unknown:${dotted}`, `Unknown key \`${dotted}\` ignored. Per-pass entries support ` +
-                "`engine`, `model`, `timeoutMs`, `enabled`, `llm`, `graphExtractionBatchSize`, " +
-                "`graphExtractionIncludeTypes`, and `lazyGraphExtraction`.");
-            cleaned ??= { ...obj };
-            delete cleaned[key];
-        }
     }
     return cleaned ?? raw;
 }, z
@@ -81,7 +63,6 @@ export const IndexPassConfigSchema = z.preprocess((raw, ctx) => {
     graphExtractionBatchSize: positiveInt.optional(),
     // Accept-any until Chunk 2 (WI-9.6c) — no longer enum-restricted.
     graphExtractionIncludeTypes: z.array(z.string().min(1)).nonempty().optional(),
-    lazyGraphExtraction: z.boolean().optional(),
 })
     .passthrough());
 const MetadataEnhanceSchema = z.object({ enabled: z.boolean().optional() }).passthrough();

package/dist/indexer/db/graph-db.js

@@ -4,7 +4,9 @@
 import { rethrowIfDataDirUnreadable, rethrowIfTestIsolationError } from "../../core/errors.js";
 import { isPathAbsent } from "../../core/path-access.js";
 import { getDbPath } from "../../core/paths.js";
+import { normalizeEntityKey } from "../../llm/graph-extract.js";
 import { closeDatabase, openExistingDatabase } from "../../storage/repositories/index-connection.js";
+import { GRAPH_SCHEMA_VERSION } from "../../storage/repositories/index-schema.js";
 function withReadableGraphDb(db, fn) {
     if (db)
         return fn(db);
@@ -26,9 +28,6 @@ function withReadableGraphDb(db, fn) {
 function uniqueSorted(values) {
     return [...new Set(values)].sort((a, b) => a.localeCompare(b));
 }
-function normalizeEntity(value) {
-    return value.trim().toLowerCase();
-}
 /** Child rows joined to their file row, the same join the loaders read through. */
 const STORED_ENTITIES = `graph_file_entities e
   JOIN graph_files gf ON gf.stash_root = e.stash_root AND gf.file_path = e.file_path AND gf.body_hash = e.body_hash
@@ -113,9 +112,9 @@ function readStoredGraphQuality(db, stashRoot) {
  * no entry_id resolution and no orphan-skip — a graph file no longer needs a
  * matching entries row.
  *
- * graph_meta records the snapshot's schema version, time and run telemetry;
- * its counts are derived from the rows as stored after the write, never from
- * the caller's in-memory graph.
+ * graph_meta records the snapshot's time and run telemetry; its counts are
+ * derived from the rows as stored after the write, never from the caller's
+ * in-memory graph.
  */
 export function replaceStoredGraph(db, graph) {
     const upsertMeta = db.prepare(`INSERT INTO graph_meta (
@@ -215,10 +214,10 @@ export function replaceStoredGraph(db, graph) {
                 insertFile.run(graph.stashRoot, node.path, fileOrder, node.type, bodyHash, node.confidence ?? null, status, reason, runId);
             }
             for (const [entityOrder, entity] of node.entities.entries()) {
-                insertEntity.run(graph.stashRoot, node.path, bodyHash, entityOrder, normalizeEntity(entity), entity);
+                insertEntity.run(graph.stashRoot, node.path, bodyHash, entityOrder, normalizeEntityKey(entity), entity);
             }
             for (const [relationOrder, relation] of node.relations.entries()) {
-                insertRelation.run(graph.stashRoot, node.path, bodyHash, relationOrder, normalizeEntity(relation.from), relation.from, normalizeEntity(relation.to), relation.to, relation.type ?? null, relation.confidence ?? null);
+                insertRelation.run(graph.stashRoot, node.path, bodyHash, relationOrder, normalizeEntityKey(relation.from), relation.from, normalizeEntityKey(relation.to), relation.to, relation.type ?? null, relation.confidence ?? null);
             }
         }
         // Delete files present in DB but absent from the new snapshot. Child
@@ -231,7 +230,7 @@ export function replaceStoredGraph(db, graph) {
             }
         }
         const quality = readStoredGraphQuality(db, graph.stashRoot);
-        upsertMeta.run(graph.stashRoot, graph.schemaVersion, graph.generatedAt, quality.consideredFiles, quality.extractedFiles, quality.entityCount, quality.relationCount, quality.extractionCoverage, quality.density, telemetry?.extractorId ?? null, telemetry?.extractionRunId ?? null, telemetry?.model ?? null, telemetry?.promptVersion ?? null, telemetry?.batchSize ?? null, telemetry?.cacheHits ?? 0, telemetry?.cacheMisses ?? 0, telemetry?.truncationCount ?? 0, telemetry?.failureCount ?? 0);
+        upsertMeta.run(graph.stashRoot, GRAPH_SCHEMA_VERSION, graph.generatedAt, quality.consideredFiles, quality.extractedFiles, quality.entityCount, quality.relationCount, quality.extractionCoverage, quality.density, telemetry?.extractorId ?? null, telemetry?.extractionRunId ?? null, telemetry?.model ?? null, telemetry?.promptVersion ?? null, telemetry?.batchSize ?? null, telemetry?.cacheHits ?? 0, telemetry?.cacheMisses ?? 0, telemetry?.truncationCount ?? 0, telemetry?.failureCount ?? 0);
     })();
 }
 export function deleteStoredGraph(db, stashPath) {
@@ -243,75 +242,6 @@ export function deleteStoredGraph(db, stashPath) {
         db.prepare("DELETE FROM graph_meta WHERE stash_root = ?").run(stashPath);
     })();
 }
-/**
- * #624-P1 — does any graph data exist for a file_path under a stash root?
- * Consumed by show/curate flows (P3) but defined here so the schema and its
- * accessors land together.
- */
-export function hasGraphData(db, stashRoot, filePath) {
-    try {
-        const row = db
-            .prepare("SELECT 1 AS present FROM graph_files WHERE stash_root = ? AND file_path = ? LIMIT 1")
-            .get(stashRoot, filePath);
-        return row !== undefined;
-    }
-    catch {
-        return false;
-    }
-}
-/**
- * #624-P3 — enqueue a file for lazy graph extraction. Idempotent on the
- * (stash_root, file_path) PK: a second enqueue refreshes body_hash + queued_at
- * and keeps the HIGHER priority. Non-blocking, no LLM call — the queued row is
- * drained later by the graph-extraction pass. Tolerant of a missing table /
- * db error (best-effort), but never masks the bun-test isolation guard.
- */
-export function enqueueGraphExtraction(db, stashRoot, filePath, bodyHash, priority = 0) {
-    try {
-        db.prepare(`INSERT INTO graph_extraction_queue (stash_root, file_path, body_hash, priority)
-         VALUES (?, ?, ?, ?)
-       ON CONFLICT(stash_root, file_path) DO UPDATE SET
-         body_hash = excluded.body_hash,
-         priority  = MAX(graph_extraction_queue.priority, excluded.priority),
-         queued_at = datetime('now')`).run(stashRoot, filePath, bodyHash, priority);
-    }
-    catch (err) {
-        rethrowIfTestIsolationError(err);
-    }
-}
-/** Read queued graph work without claiming or deleting it. */
-export function peekExtractionQueue(db, stashRoot, limit) {
-    try {
-        const rows = db
-            .prepare(`SELECT file_path, body_hash, priority
-           FROM graph_extraction_queue
-           WHERE stash_root = ?
-           ORDER BY priority DESC, queued_at ASC
-           LIMIT ?`)
-            .all(stashRoot, limit);
-        return rows.map((row) => ({ filePath: row.file_path, bodyHash: row.body_hash, priority: row.priority }));
-    }
-    catch (err) {
-        rethrowIfTestIsolationError(err);
-        return [];
-    }
-}
-/**
- * Acknowledge the exact queued revision that was classified and handled.
- * A concurrent re-enqueue with a new body hash therefore survives.
- */
-export function acknowledgeExtractionQueueEntry(db, stashRoot, filePath, bodyHash) {
-    try {
-        const result = db
-            .prepare("DELETE FROM graph_extraction_queue WHERE stash_root = ? AND file_path = ? AND body_hash = ?")
-            .run(stashRoot, filePath, bodyHash);
-        return result.changes > 0;
-    }
-    catch (err) {
-        rethrowIfTestIsolationError(err);
-        return false;
-    }
-}
 /**
  * Scoped loader — graph_files rows without entities/relations. Used for
  * orphan detection and entity overview commands.
@@ -350,7 +280,6 @@ export function loadStoredGraphMeta(stashPath, db) {
             const row = readDb
                 .prepare(`SELECT
              stash_root,
-             schema_version,
              generated_at,
              considered_files,
              extracted_files,
@@ -375,7 +304,6 @@ export function loadStoredGraphMeta(stashPath, db) {
             return {
                 stashPath: row.stash_root,
                 graphPath: getDbPath(),
-                schemaVersion: row.schema_version,
                 generatedAt: row.generated_at,
                 quality: {
                     consideredFiles: row.considered_files,
@@ -484,7 +412,6 @@ export function loadStoredGraphSnapshot(stashPath, db) {
             return {
                 stashPath: meta.stashPath,
                 graphPath: meta.graphPath,
-                schemaVersion: meta.schemaVersion,
                 generatedAt: meta.generatedAt,
                 ...(meta.quality ? { quality: meta.quality } : {}),
                 ...(meta.telemetry ? { telemetry: meta.telemetry } : {}),