recallnest 2.2.0 → 2.2.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (232) hide show
  1. package/.env +4 -0
  2. package/config.json +55 -0
  3. package/logs/daily-recall.error.log +0 -0
  4. package/logs/daily-recall.log +30 -0
  5. package/logs/ingest-2026-04-11.log +110 -0
  6. package/logs/ingest-2026-04-12.log +90 -0
  7. package/logs/launchagent-stderr.log +0 -0
  8. package/logs/launchagent-stdout.log +0 -0
  9. package/package.json +1 -1
  10. package/.claude-plugin/CLAUDE.md +0 -29
  11. package/.claude-plugin/marketplace.json +0 -18
  12. package/.dockerignore +0 -11
  13. package/.github/workflows/ci.yml +0 -28
  14. package/.mcp.json +0 -9
  15. package/CHANGELOG.md +0 -51
  16. package/DEMO.md +0 -94
  17. package/Dockerfile +0 -20
  18. package/EVAL.md +0 -89
  19. package/FAILURES.md +0 -75
  20. package/OPERATIONS.md +0 -59
  21. package/ROADMAP.md +0 -245
  22. package/SKILL.md +0 -250
  23. package/assets/dashboard.png +0 -0
  24. package/assets/knowledge-graph.png +0 -0
  25. package/assets/recallnest-workbench.png +0 -0
  26. package/assets/screenshots/ui-full.png +0 -0
  27. package/assets/screenshots/ui-overview.png +0 -0
  28. package/assets/screenshots/ui-search.png +0 -0
  29. package/assets/ui/app.js +0 -1031
  30. package/assets/ui/index.html +0 -224
  31. package/assets/ui/styles.css +0 -781
  32. package/assets/wechat-qr.jpg +0 -0
  33. package/bun.lock +0 -337
  34. package/daily-reflections/2026-03-05.md +0 -15
  35. package/daily-reflections/2026-03-06.md +0 -15
  36. package/daily-reflections/2026-03-07.md +0 -7
  37. package/daily-reflections/2026-03-08.md +0 -34
  38. package/daily-reflections/2026-03-09.md +0 -35
  39. package/daily-reflections/2026-03-10.md +0 -36
  40. package/daily-reflections/2026-03-11.md +0 -35
  41. package/daily-reflections/2026-03-12.md +0 -36
  42. package/daily-reflections/2026-03-13.md +0 -35
  43. package/daily-reflections/2026-03-14.md +0 -35
  44. package/daily-reflections/2026-03-15.md +0 -35
  45. package/daily-reflections/2026-03-16.md +0 -41
  46. package/daily-reflections/2026-03-17.md +0 -10
  47. package/daily-reflections/2026-03-18.md +0 -10
  48. package/daily-reflections/2026-03-19.md +0 -10
  49. package/daily-reflections/2026-03-20.md +0 -10
  50. package/daily-reflections/2026-03-21.md +0 -10
  51. package/daily-reflections/2026-03-22.md +0 -10
  52. package/daily-reflections/2026-03-23.md +0 -10
  53. package/daily-reflections/2026-03-24.md +0 -10
  54. package/daily-reflections/2026-03-25.md +0 -32
  55. package/daily-reflections/2026-03-26.md +0 -37
  56. package/daily-reflections/2026-03-27.md +0 -35
  57. package/daily-reflections/2026-03-28.md +0 -35
  58. package/daily-reflections/2026-03-29.md +0 -35
  59. package/daily-reflections/2026-03-30.md +0 -36
  60. package/daily-reflections/2026-03-31.md +0 -35
  61. package/daily-reflections/2026-04-01.md +0 -35
  62. package/daily-reflections/2026-04-02.md +0 -35
  63. package/daily-reflections/2026-04-03.md +0 -37
  64. package/daily-reflections/2026-04-04.md +0 -37
  65. package/daily-reflections/2026-04-05.md +0 -35
  66. package/daily-reflections/2026-04-06.md +0 -36
  67. package/daily-reflections/2026-04-07.md +0 -35
  68. package/daily-reflections/2026-04-08.md +0 -35
  69. package/daily-reflections/2026-04-12.md +0 -35
  70. package/data/alias-map.example.json +0 -22
  71. package/docs/api-reference.md +0 -1291
  72. package/docs/architecture.md +0 -168
  73. package/docs/memory-boundary-contract.md +0 -226
  74. package/docs/memory-categories.md +0 -119
  75. package/docs/pipeline-audit-ultramemory.md +0 -177
  76. package/docs/plans/2026-03-08-query-expansion.md +0 -270
  77. package/docs/plans/2026-03-16-continuity-layer.md +0 -196
  78. package/docs/plans/2026-04-01-recall-boost.md +0 -100
  79. package/docs/plans/mcp-tiering-plan.md +0 -126
  80. package/docs/research/recall-quality-audit-2026-03-22.md +0 -153
  81. package/docs/self-evolution.md +0 -198
  82. package/docs/superpowers/plans/2026-03-19-ingest-pipeline-upgrade.md +0 -782
  83. package/docs/superpowers/plans/2026-04-11-constructive-retrieval-emotion.md +0 -1195
  84. package/docs/superpowers/specs/2026-03-19-ingest-pipeline-upgrade-design.md +0 -77
  85. package/docs/superpowers/specs/2026-03-23-pull-claude-desktop-design.md +0 -95
  86. package/docs/superpowers/specs/2026-04-11-constructive-retrieval-emotion-design.md +0 -398
  87. package/eval/continuity/results-2026-03-16.md +0 -117
  88. package/eval/continuity/results-2026-03-19.md +0 -453
  89. package/eval/continuity/runbook.md +0 -110
  90. package/eval/continuity/seed-cases.md +0 -54
  91. package/eval/reports/2026-03-06-abstract-query-baseline.md +0 -96
  92. package/eval/reports/2026-03-06-aws-ssh-baseline.md +0 -82
  93. package/eval/reports/2026-03-06-baseline.md +0 -82
  94. package/eval/reports/query-expansion.md +0 -138
  95. package/glama.json +0 -6
  96. package/lm +0 -19
  97. package/plan-philosophy-upgrades.md +0 -363
  98. package/plan-product-upgrades.md +0 -155
  99. package/plan-v1.2-distributable.md +0 -105
  100. package/plan.md +0 -140
  101. package/research-memory-evolution.md +0 -440
  102. package/research-rl-agent-memory.md +0 -330
  103. package/scripts/archive-distilled-junk.ts +0 -116
  104. package/scripts/backfill-anchors.ts +0 -130
  105. package/scripts/backfill-l0l1l2.ts +0 -98
  106. package/scripts/blind-scan.ts +0 -588
  107. package/scripts/cleanup-step1-exact-dups.ts +0 -117
  108. package/scripts/cleanup-step2-template-dups.ts +0 -113
  109. package/scripts/cleanup-step3-stale-content.ts +0 -149
  110. package/scripts/cleanup-step4-orphan-scopes.ts +0 -131
  111. package/scripts/delete-distilled-junk.ts +0 -88
  112. package/scripts/diagnose-vector-search.ts +0 -132
  113. package/scripts/distill-facts.ts +0 -305
  114. package/scripts/health-check.ts +0 -343
  115. package/scripts/incremental-ingest.sh +0 -71
  116. package/scripts/migrate-language.ts +0 -75
  117. package/scripts/mlp-health-check.ts +0 -382
  118. package/scripts/pull-claude-desktop.ts +0 -515
  119. package/scripts/re-ingest-cc.sh +0 -66
  120. package/scripts/rebalance-memory-health.ts +0 -181
  121. package/scripts/run-desktop-ingest.sh +0 -11
  122. package/scripts/start-server.sh +0 -12
  123. package/scripts/verify-emotion-decay.ts +0 -148
  124. package/scripts/viz-dashboard.html +0 -423
  125. package/scripts/viz-export.ts +0 -703
  126. package/src/__tests__/access-tracker-novelty.test.ts +0 -228
  127. package/src/__tests__/activity-counter.test.ts +0 -96
  128. package/src/__tests__/admission-control.test.ts +0 -206
  129. package/src/__tests__/anchor-generator.test.ts +0 -67
  130. package/src/__tests__/auto-consolidation.test.ts +0 -168
  131. package/src/__tests__/auto-recall.test.ts +0 -131
  132. package/src/__tests__/babel-integration.test.ts +0 -139
  133. package/src/__tests__/batch-dedup.test.ts +0 -65
  134. package/src/__tests__/capture-engine.test.ts +0 -1144
  135. package/src/__tests__/capture-heuristic.test.ts +0 -171
  136. package/src/__tests__/cascade-forget.test.ts +0 -210
  137. package/src/__tests__/category-min-scores.test.ts +0 -33
  138. package/src/__tests__/cc8-cc1-essential-context.test.ts +0 -298
  139. package/src/__tests__/circuit-breaker.test.ts +0 -99
  140. package/src/__tests__/cluster-consolidation.test.ts +0 -712
  141. package/src/__tests__/confidence-metadata.test.ts +0 -125
  142. package/src/__tests__/confidence-tracker.test.ts +0 -249
  143. package/src/__tests__/conflict-advisor.test.ts +0 -206
  144. package/src/__tests__/conflict-escalation.test.ts +0 -149
  145. package/src/__tests__/conflict-lifecycle.test.ts +0 -83
  146. package/src/__tests__/conflict-output.test.ts +0 -85
  147. package/src/__tests__/conflict-store.test.ts +0 -374
  148. package/src/__tests__/consolidation-engine.test.ts +0 -181
  149. package/src/__tests__/constructive-retrieval.test.ts +0 -509
  150. package/src/__tests__/context-collapse-renderer.test.ts +0 -214
  151. package/src/__tests__/context-composer-pins.test.ts +0 -64
  152. package/src/__tests__/context-composer-task-selection.test.ts +0 -24
  153. package/src/__tests__/context-reconstructor.test.ts +0 -65
  154. package/src/__tests__/context-renderer.test.ts +0 -91
  155. package/src/__tests__/continuity-eval.test.ts +0 -275
  156. package/src/__tests__/continuity-schema.test.ts +0 -97
  157. package/src/__tests__/conversation-importer.test.ts +0 -692
  158. package/src/__tests__/core-recall-eval.test.ts +0 -235
  159. package/src/__tests__/core-summary.test.ts +0 -109
  160. package/src/__tests__/data-checkup.test.ts +0 -225
  161. package/src/__tests__/decay-exemption.test.ts +0 -84
  162. package/src/__tests__/dedup-actions.test.ts +0 -58
  163. package/src/__tests__/dedup-guard.test.ts +0 -39
  164. package/src/__tests__/deterministic-id.test.ts +0 -39
  165. package/src/__tests__/distill-lock.test.ts +0 -176
  166. package/src/__tests__/doctor.test.ts +0 -303
  167. package/src/__tests__/dream-pipeline.test.ts +0 -232
  168. package/src/__tests__/drill-down.test.ts +0 -108
  169. package/src/__tests__/embedder.test.ts +0 -122
  170. package/src/__tests__/emotion-decay.test.ts +0 -46
  171. package/src/__tests__/emotion-detector.test.ts +0 -56
  172. package/src/__tests__/emotion-phase1-integration.test.ts +0 -153
  173. package/src/__tests__/emotion-retrieval.test.ts +0 -56
  174. package/src/__tests__/entity-resolver.test.ts +0 -58
  175. package/src/__tests__/ethics-layer.test.ts +0 -476
  176. package/src/__tests__/event-segmenter.test.ts +0 -153
  177. package/src/__tests__/feature-flags-integration.test.ts +0 -54
  178. package/src/__tests__/frequency-tracker.test.ts +0 -113
  179. package/src/__tests__/graph-export.test.ts +0 -489
  180. package/src/__tests__/hotness-blend.test.ts +0 -76
  181. package/src/__tests__/ingest.test.ts +0 -360
  182. package/src/__tests__/interference-detector.test.ts +0 -203
  183. package/src/__tests__/kg-extractor.test.ts +0 -184
  184. package/src/__tests__/kg-store.test.ts +0 -198
  185. package/src/__tests__/llm-consolidation.test.ts +0 -226
  186. package/src/__tests__/mcp-efficiency.test.ts +0 -288
  187. package/src/__tests__/memory-boundaries.test.ts +0 -237
  188. package/src/__tests__/memory-evolution-b2-b3-e1.test.ts +0 -243
  189. package/src/__tests__/memory-evolution.test.ts +0 -271
  190. package/src/__tests__/memory-health-rebalance.test.ts +0 -144
  191. package/src/__tests__/memory-lint.test.ts +0 -271
  192. package/src/__tests__/memory-output.test.ts +0 -215
  193. package/src/__tests__/multi-vector.test.ts +0 -242
  194. package/src/__tests__/narrative-architecture.test.ts +0 -456
  195. package/src/__tests__/pending-review.test.ts +0 -77
  196. package/src/__tests__/post-retrieval-filter.test.ts +0 -178
  197. package/src/__tests__/ppr-traversal.test.ts +0 -256
  198. package/src/__tests__/prediction-engine.test.ts +0 -349
  199. package/src/__tests__/predictive-prospective-memory.test.ts +0 -416
  200. package/src/__tests__/preference-guard-smoke.test.ts +0 -211
  201. package/src/__tests__/preference-matcher.test.ts +0 -230
  202. package/src/__tests__/preference-slots.test.ts +0 -205
  203. package/src/__tests__/prospective-memory.test.ts +0 -283
  204. package/src/__tests__/query-entity-detector.test.ts +0 -171
  205. package/src/__tests__/query-expander-alias.test.ts +0 -89
  206. package/src/__tests__/query-expander.test.ts +0 -34
  207. package/src/__tests__/recall-governor.test.ts +0 -227
  208. package/src/__tests__/recall-mode.test.ts +0 -64
  209. package/src/__tests__/result-synthesizer.test.ts +0 -198
  210. package/src/__tests__/resume-context.test.ts +0 -4898
  211. package/src/__tests__/retrieval-trace.test.ts +0 -213
  212. package/src/__tests__/retriever-boundary-weight.test.ts +0 -136
  213. package/src/__tests__/retroactive-boost.test.ts +0 -223
  214. package/src/__tests__/salience-scoring.test.ts +0 -106
  215. package/src/__tests__/scope-inventory.test.ts +0 -200
  216. package/src/__tests__/scope-policy.test.ts +0 -69
  217. package/src/__tests__/session-checkpoint.test.ts +0 -259
  218. package/src/__tests__/session-distiller.test.ts +0 -611
  219. package/src/__tests__/session-gc.test.ts +0 -215
  220. package/src/__tests__/skill-engine.test.ts +0 -546
  221. package/src/__tests__/skill-promotion.test.ts +0 -349
  222. package/src/__tests__/store-scope-audit.test.ts +0 -81
  223. package/src/__tests__/supersede-chain.test.ts +0 -154
  224. package/src/__tests__/temporal-parser.test.ts +0 -241
  225. package/src/__tests__/temporal-validity.test.ts +0 -160
  226. package/src/__tests__/term-registry.test.ts +0 -121
  227. package/src/__tests__/tool-output-compressor.test.ts +0 -133
  228. package/src/__tests__/topic-tag.test.ts +0 -113
  229. package/src/__tests__/trusted-memory.test.ts +0 -828
  230. package/src/__tests__/version-manager.test.ts +0 -213
  231. package/src/__tests__/workflow-observation.test.ts +0 -400
  232. package/src/__tests__/write-verifier.test.ts +0 -122
@@ -1,213 +0,0 @@
1
- /**
2
- * Tests for Tier 3.3: Version Coexistence + Competition
3
- *
4
- * Validates:
5
- * 1. computeVersionRank formula
6
- * 2. createVersionGroup tags both entries with same group
7
- * 3. deduplicateByVersionGroup keeps top-ranked per group
8
- * 4. Entries without version_group pass through unchanged
9
- * 5. Multiple version groups are handled independently
10
- * 6. Existing version_group is reused (not duplicated)
11
- */
12
- import { describe, expect, it } from "bun:test";
13
- import {
14
- computeVersionRank,
15
- deduplicateByVersionGroup,
16
- createVersionGroup,
17
- } from "../version-manager.js";
18
- import type { MemoryEntry } from "../store.js";
19
-
20
- // ---------------------------------------------------------------------------
21
- // Helpers
22
- // ---------------------------------------------------------------------------
23
-
24
- function makeEntry(
25
- id: string,
26
- overrides: {
27
- confidence?: number;
28
- accessCount?: number;
29
- version_group?: string;
30
- version_rank?: number;
31
- } = {},
32
- ): MemoryEntry {
33
- const meta: Record<string, unknown> = {
34
- confidence: overrides.confidence ?? 0.7,
35
- accessCount: overrides.accessCount ?? 0,
36
- };
37
- if (overrides.version_group) meta.version_group = overrides.version_group;
38
- if (overrides.version_rank !== undefined) meta.version_rank = overrides.version_rank;
39
-
40
- return {
41
- id,
42
- text: `memory-${id}`,
43
- vector: [1, 0, 0],
44
- category: "events",
45
- scope: "project:test",
46
- importance: 0.7,
47
- timestamp: Date.now(),
48
- metadata: JSON.stringify(meta),
49
- };
50
- }
51
-
52
- function createMockStore(entries: MemoryEntry[]) {
53
- const data = new Map(entries.map(e => [e.id, { ...e }]));
54
- const updates: Array<{ id: string; metadata: string }> = [];
55
-
56
- return {
57
- data,
58
- updates,
59
- async getById(id: string) {
60
- return data.get(id) ?? null;
61
- },
62
- async update(id: string, upd: { metadata?: string }, _scope?: string[]) {
63
- const entry = data.get(id);
64
- if (!entry) return null;
65
- if (upd.metadata) {
66
- entry.metadata = upd.metadata;
67
- updates.push({ id, metadata: upd.metadata });
68
- }
69
- return entry;
70
- },
71
- };
72
- }
73
-
74
- // ---------------------------------------------------------------------------
75
- // computeVersionRank
76
- // ---------------------------------------------------------------------------
77
-
78
- describe("computeVersionRank", () => {
79
- it("computes rank from importance, confidence, and accessCount", () => {
80
- const entry = makeEntry("a", { confidence: 1.0, accessCount: 10 });
81
- const rank = computeVersionRank(entry);
82
- // quality = 0.5*0.7 + 0.5*1.0 = 0.85 (importance=0.7 from makeEntry)
83
- // rank = 0.85 * (1 + ln(11)) ≈ 0.85 * 3.398 = 2.888
84
- expect(rank).toBeGreaterThan(2.8);
85
- expect(rank).toBeLessThan(3.0);
86
- });
87
-
88
- it("returns baseline for default confidence and zero access", () => {
89
- const entry = makeEntry("b", { confidence: 0.7, accessCount: 0 });
90
- const rank = computeVersionRank(entry);
91
- // quality = 0.5*0.7 + 0.5*0.7 = 0.7
92
- // rank = 0.7 * (1 + 0) = 0.7
93
- expect(rank).toBeCloseTo(0.7, 5);
94
- });
95
-
96
- it("higher confidence wins over lower with same access", () => {
97
- const high = computeVersionRank(makeEntry("h", { confidence: 1.0, accessCount: 5 }));
98
- const low = computeVersionRank(makeEntry("l", { confidence: 0.3, accessCount: 5 }));
99
- expect(high).toBeGreaterThan(low);
100
- });
101
-
102
- it("higher access wins over lower with same confidence", () => {
103
- const frequent = computeVersionRank(makeEntry("f", { confidence: 0.7, accessCount: 20 }));
104
- const rare = computeVersionRank(makeEntry("r", { confidence: 0.7, accessCount: 1 }));
105
- expect(frequent).toBeGreaterThan(rare);
106
- });
107
- });
108
-
109
- // ---------------------------------------------------------------------------
110
- // createVersionGroup
111
- // ---------------------------------------------------------------------------
112
-
113
- describe("createVersionGroup", () => {
114
- it("tags both entries with the same version_group", async () => {
115
- const canonical = makeEntry("canon", { confidence: 1.0, accessCount: 5 });
116
- const member = makeEntry("member", { confidence: 0.5, accessCount: 1 });
117
- const store = createMockStore([canonical, member]);
118
-
119
- const groupId = await createVersionGroup(store as any, canonical, member, "project:test");
120
-
121
- expect(groupId).toBeTruthy();
122
- expect(store.updates.length).toBe(2);
123
-
124
- const canonMeta = JSON.parse(store.data.get("canon")!.metadata);
125
- const memberMeta = JSON.parse(store.data.get("member")!.metadata);
126
-
127
- expect(canonMeta.version_group).toBe(groupId);
128
- expect(memberMeta.version_group).toBe(groupId);
129
- expect(canonMeta.version_rank).toBeGreaterThan(memberMeta.version_rank);
130
- });
131
-
132
- it("reuses existing version_group from canonical", async () => {
133
- const canonical = makeEntry("canon", {
134
- confidence: 1.0,
135
- accessCount: 5,
136
- version_group: "vg-existing",
137
- version_rank: 3.0,
138
- });
139
- const member = makeEntry("new", { confidence: 0.5, accessCount: 0 });
140
- const store = createMockStore([canonical, member]);
141
-
142
- const groupId = await createVersionGroup(store as any, canonical, member, "project:test");
143
-
144
- expect(groupId).toBe("vg-existing");
145
- const memberMeta = JSON.parse(store.data.get("new")!.metadata);
146
- expect(memberMeta.version_group).toBe("vg-existing");
147
- });
148
- });
149
-
150
- // ---------------------------------------------------------------------------
151
- // deduplicateByVersionGroup
152
- // ---------------------------------------------------------------------------
153
-
154
- describe("deduplicateByVersionGroup", () => {
155
- it("keeps only the top-ranked entry per version group", () => {
156
- const results = [
157
- { entry: makeEntry("a", { version_group: "vg1", version_rank: 3.0 }), score: 0.9, sources: {} },
158
- { entry: makeEntry("b", { version_group: "vg1", version_rank: 1.0 }), score: 0.85, sources: {} },
159
- { entry: makeEntry("c"), score: 0.8, sources: {} },
160
- ];
161
-
162
- const deduped = deduplicateByVersionGroup(results);
163
-
164
- expect(deduped.length).toBe(2);
165
- expect(deduped.map(r => r.entry.id)).toContain("a");
166
- expect(deduped.map(r => r.entry.id)).toContain("c");
167
- expect(deduped.map(r => r.entry.id)).not.toContain("b");
168
- });
169
-
170
- it("passes through entries without version_group", () => {
171
- const results = [
172
- { entry: makeEntry("x"), score: 0.9, sources: {} },
173
- { entry: makeEntry("y"), score: 0.8, sources: {} },
174
- ];
175
-
176
- const deduped = deduplicateByVersionGroup(results);
177
-
178
- expect(deduped.length).toBe(2);
179
- });
180
-
181
- it("handles multiple independent version groups", () => {
182
- const results = [
183
- { entry: makeEntry("a1", { version_group: "vg1", version_rank: 3.0 }), score: 0.95, sources: {} },
184
- { entry: makeEntry("a2", { version_group: "vg1", version_rank: 1.0 }), score: 0.90, sources: {} },
185
- { entry: makeEntry("b1", { version_group: "vg2", version_rank: 2.0 }), score: 0.85, sources: {} },
186
- { entry: makeEntry("b2", { version_group: "vg2", version_rank: 4.0 }), score: 0.80, sources: {} },
187
- { entry: makeEntry("solo"), score: 0.75, sources: {} },
188
- ];
189
-
190
- const deduped = deduplicateByVersionGroup(results);
191
-
192
- expect(deduped.length).toBe(3);
193
- const ids = deduped.map(r => r.entry.id);
194
- expect(ids).toContain("a1"); // vg1 winner (rank 3.0)
195
- expect(ids).toContain("b2"); // vg2 winner (rank 4.0)
196
- expect(ids).toContain("solo");
197
- });
198
-
199
- it("handles empty array", () => {
200
- expect(deduplicateByVersionGroup([])).toEqual([]);
201
- });
202
-
203
- it("handles single entry in a version group", () => {
204
- const results = [
205
- { entry: makeEntry("only", { version_group: "vg1", version_rank: 2.0 }), score: 0.9, sources: {} },
206
- ];
207
-
208
- const deduped = deduplicateByVersionGroup(results);
209
-
210
- expect(deduped.length).toBe(1);
211
- expect(deduped[0].entry.id).toBe("only");
212
- });
213
- });
@@ -1,400 +0,0 @@
1
- import { describe, expect, it } from "bun:test";
2
- import { mkdirSync, mkdtempSync, readFileSync, rmSync, writeFileSync } from "node:fs";
3
- import { tmpdir } from "node:os";
4
- import { join } from "node:path";
5
-
6
- import {
7
- applyWorkflowObservationScopeSuggestions,
8
- formatWorkflowObservationScopeReview,
9
- keepWorkflowObservationScopesGlobal,
10
- parseWorkflowObservationScopeCue,
11
- reviewWorkflowObservationScopes,
12
- } from "../workflow-observation-scope-review.js";
13
- import {
14
- buildWorkflowEvidence,
15
- buildWorkflowHealthDashboard,
16
- buildWorkflowObservationRecord,
17
- inspectWorkflowHealth,
18
- resolveWorkflowObservationScope,
19
- } from "../workflow-observation-engine.js";
20
- import { collectScopeInventory } from "../scope-inventory.js";
21
- import { buildSessionCheckpointResult } from "../session-engine.js";
22
- import { MemoryStore } from "../store.js";
23
- import { buildManagedCheckpointObservation, buildManagedResumeObservation } from "../workflow-observation-managed.js";
24
- import { WorkflowObservationStore } from "../workflow-observation-store.js";
25
-
26
- describe("workflow observation engine", () => {
27
- it("defaults observation scope to global", () => {
28
- const record = buildWorkflowObservationRecord({
29
- workflowId: "resume_context",
30
- outcome: "missed",
31
- summary: "Fresh window skipped resume_context before repo exploration.",
32
- });
33
-
34
- expect(resolveWorkflowObservationScope(record)).toBe("global");
35
- expect(record.resolvedScope).toBe("global");
36
- });
37
-
38
- it("aggregates workflow health and evidence from append-only observations", async () => {
39
- const dir = mkdtempSync(join(tmpdir(), "recallnest-workflow-observations-"));
40
- try {
41
- const store = new WorkflowObservationStore(dir);
42
- const records = [
43
- {
44
- workflowId: "resume_context",
45
- scope: "project:recallnest",
46
- outcome: "missed",
47
- summary: "Fresh window explored the repo before continuity recovery.",
48
- signal: "missed-startup-trigger",
49
- source: "smoke",
50
- recordedAt: "2026-03-15T03:00:00.000Z",
51
- },
52
- {
53
- workflowId: "resume_context",
54
- scope: "project:recallnest",
55
- outcome: "corrected",
56
- summary: "User had to remind the agent to recover continuity first.",
57
- signal: "user-correction",
58
- source: "agent",
59
- recordedAt: "2026-03-16T03:00:00.000Z",
60
- },
61
- {
62
- workflowId: "resume_context",
63
- scope: "project:recallnest",
64
- outcome: "success",
65
- summary: "Fresh window recovered RecallNest continuity before coding.",
66
- signal: "startup-recovered",
67
- source: "smoke",
68
- recordedAt: "2026-03-17T03:00:00.000Z",
69
- },
70
- {
71
- workflowId: "checkpoint_session",
72
- scope: "project:recallnest",
73
- outcome: "failure",
74
- summary: "Checkpoint still carried repo-state text before the product-side guard landed.",
75
- signal: "repo-state-contamination",
76
- source: "smoke",
77
- recordedAt: "2026-03-17T04:00:00.000Z",
78
- },
79
- ];
80
-
81
- for (const record of records) {
82
- await store.save(buildWorkflowObservationRecord(record));
83
- }
84
-
85
- const health = await inspectWorkflowHealth(store, {
86
- workflowId: "resume_context",
87
- scope: "project:recallnest",
88
- now: new Date("2026-03-17T12:00:00.000Z"),
89
- });
90
- expect(health.status).toBe("watch");
91
- expect(health.windows[1]?.total).toBe(3);
92
- expect(health.windows[1]?.missed).toBe(1);
93
- expect(health.windows[1]?.corrected).toBe(1);
94
- expect(health.windows[1]?.successRate).toBeCloseTo(1 / 3, 5);
95
-
96
- const dashboard = buildWorkflowHealthDashboard(
97
- await store.listRecent({ scope: "project:recallnest", limit: 50 }),
98
- { scope: "project:recallnest" },
99
- );
100
- expect(dashboard[0]?.workflowId).toBe("checkpoint_session");
101
- expect(dashboard[0]?.status).toBe("critical");
102
-
103
- const evidence = await buildWorkflowEvidence(store, {
104
- workflowId: "checkpoint_session",
105
- scope: "project:recallnest",
106
- now: new Date("2026-03-17T12:00:00.000Z"),
107
- });
108
- expect(evidence.topSignals[0]?.signal).toBe("repo-state-contamination");
109
- expect(evidence.suggestions).toContain(
110
- "Keep volatile repo-state text out of saved checkpoints and handoff summaries unless this window verified it.",
111
- );
112
- expect(evidence.suggestions).toContain(
113
- "Add end-of-window guards so checkpoint content is sanitized before it becomes the next handoff.",
114
- );
115
- } finally {
116
- rmSync(dir, { recursive: true, force: true });
117
- }
118
- });
119
-
120
- it("builds managed continuity observations without routing them into regular memory", () => {
121
- const resumeObservation = buildManagedResumeObservation({
122
- sessionId: "session-123",
123
- task: "Continue RecallNest continuity work",
124
- }, {
125
- stableContext: ["Entity: RecallNest continuity revolves around three primitives."],
126
- relevantPatterns: ["Call resume_context before coding."],
127
- recentCases: ["Case: RecallNest sparse startup context cleanup"],
128
- latestCheckpoint: {
129
- sessionId: "session-123",
130
- summary: "Checkpoint summary",
131
- updatedAt: "2026-03-17T11:00:00.000Z",
132
- },
133
- responseMode: "default",
134
- });
135
-
136
- expect(resumeObservation.workflowId).toBe("resume_context");
137
- expect(resumeObservation.outcome).toBe("success");
138
- expect(resumeObservation.scope).toBe("session:session-123");
139
- expect(resumeObservation.source).toBe("managed");
140
- expect(resumeObservation.signal).toBe("managed-resume-resolved");
141
- expect(resumeObservation.tags).toContain("managed");
142
- expect(resumeObservation.tools).toEqual(["resume_context"]);
143
-
144
- const checkpointObservation = buildManagedCheckpointObservation(buildSessionCheckpointResult({
145
- sessionId: "session-123",
146
- scope: "project:recallnest",
147
- summary: "Only resumed context here. git status shows modified files.",
148
- openLoops: ["git status still needs review"],
149
- }));
150
-
151
- expect(checkpointObservation.workflowId).toBe("checkpoint_session");
152
- expect(checkpointObservation.outcome).toBe("corrected");
153
- expect(checkpointObservation.scope).toBe("project:recallnest");
154
- expect(checkpointObservation.source).toBe("managed");
155
- expect(checkpointObservation.signal).toBe("repo-state-sanitized");
156
- expect(checkpointObservation.summary).toContain("summary and openLoops");
157
- expect(checkpointObservation.tools).toEqual(["checkpoint_session"]);
158
- });
159
-
160
- it("uses resolved scope fallback for managed resume observations when request scope is absent", () => {
161
- const observation = buildManagedResumeObservation({
162
- task: "Continue RecallNest scope cleanup",
163
- }, {
164
- resolvedScope: "project:recallnest",
165
- stableContext: ["Entity: RecallNest continuity revolves around three primitives."],
166
- relevantPatterns: [],
167
- recentCases: [],
168
- latestCheckpoint: {
169
- sessionId: "session-xyz",
170
- resolvedScope: "project:recallnest",
171
- summary: "Checkpoint summary",
172
- updatedAt: "2026-03-18T00:00:00.000Z",
173
- },
174
- responseMode: "default",
175
- });
176
-
177
- expect(observation.scope).toBe("project:recallnest");
178
- });
179
-
180
- it("reviews and applies cue-based scope suggestions for legacy global workflow observations", () => {
181
- const dir = mkdtempSync(join(tmpdir(), "recallnest-workflow-scope-review-"));
182
- try {
183
- const recallnestPath = join(dir, "recallnest.json");
184
- const externalPath = join(dir, "external.json");
185
- writeFileSync(recallnestPath, JSON.stringify({
186
- workflowId: "resume_context",
187
- summary: "Managed resume_context recovered 1 stable item(s), 1 pattern(s), and 1 case(s).",
188
- source: "managed",
189
- task: "Continue RecallNest scope cleanup",
190
- recordedAt: "2026-03-18T00:00:00.000Z",
191
- observationId: "obs-recallnest",
192
- resolvedScope: "global",
193
- }, null, 2) + "\n");
194
- writeFileSync(externalPath, JSON.stringify({
195
- workflowId: "resume_context",
196
- summary: "Managed resume_context recovered 1 stable item(s), 0 pattern(s), and 0 case(s).",
197
- source: "managed",
198
- task: "A2A code Claude SDK calling error",
199
- recordedAt: "2026-03-18T00:01:00.000Z",
200
- observationId: "obs-external",
201
- resolvedScope: "global",
202
- }, null, 2) + "\n");
203
-
204
- const review = reviewWorkflowObservationScopes({
205
- dir,
206
- currentScope: "global",
207
- source: "managed",
208
- workflowId: "resume_context",
209
- cues: [parseWorkflowObservationScopeCue("recallnest=project:recallnest")],
210
- });
211
-
212
- expect(review.totalCount).toBe(2);
213
- expect(review.suggestedCount).toBe(1);
214
- expect(review.manualReviewCount).toBe(1);
215
- expect(review.entries[0]?.suggestedScope).toBeUndefined();
216
- expect(review.entries[1]?.suggestedScope).toBe("project:recallnest");
217
- expect(formatWorkflowObservationScopeReview(review)).toContain("project:recallnest");
218
-
219
- const result = applyWorkflowObservationScopeSuggestions(review);
220
- expect(result.updatedCount).toBe(1);
221
- expect(result.skippedCount).toBe(1);
222
-
223
- const rewritten = JSON.parse(readFileSync(recallnestPath, "utf-8"));
224
- const untouched = JSON.parse(readFileSync(externalPath, "utf-8"));
225
- expect(rewritten.scope).toBe("project:recallnest");
226
- expect(rewritten.resolvedScope).toBe("project:recallnest");
227
- expect(untouched.resolvedScope).toBe("global");
228
- } finally {
229
- rmSync(dir, { recursive: true, force: true });
230
- }
231
- });
232
-
233
- it("can infer a scope from a nearby managed observation when cues are absent", () => {
234
- const dir = mkdtempSync(join(tmpdir(), "recallnest-workflow-scope-neighbor-"));
235
- try {
236
- const globalPath = join(dir, "global.json");
237
- const neighborPath = join(dir, "neighbor.json");
238
- writeFileSync(globalPath, JSON.stringify({
239
- workflowId: "resume_context",
240
- summary: "Managed resume_context recovered 2 stable item(s), 1 pattern(s), and 0 case(s).",
241
- source: "managed",
242
- recordedAt: "2026-03-18T00:00:00.000Z",
243
- observationId: "obs-global",
244
- resolvedScope: "global",
245
- }, null, 2) + "\n");
246
- writeFileSync(neighborPath, JSON.stringify({
247
- workflowId: "checkpoint_session",
248
- summary: "Managed checkpoint_session sanitized repo-state text out of summary before saving the handoff.",
249
- source: "managed",
250
- recordedAt: "2026-03-18T00:00:40.000Z",
251
- observationId: "obs-neighbor",
252
- resolvedScope: "session:recallnest-2026-03-18",
253
- }, null, 2) + "\n");
254
-
255
- const review = reviewWorkflowObservationScopes({
256
- dir,
257
- currentScope: "global",
258
- source: "managed",
259
- workflowId: "resume_context",
260
- neighborWindowSeconds: 90,
261
- });
262
-
263
- expect(review.totalCount).toBe(1);
264
- expect(review.suggestedCount).toBe(1);
265
- expect(review.entries[0]?.suggestedScope).toBe("session:recallnest-2026-03-18");
266
- expect(review.entries[0]?.reason).toContain("matched nearby checkpoint_session");
267
- } finally {
268
- rmSync(dir, { recursive: true, force: true });
269
- }
270
- });
271
-
272
- it("can mark manual-review legacy global observations as reviewed keep-global", () => {
273
- const dir = mkdtempSync(join(tmpdir(), "recallnest-workflow-scope-keep-global-"));
274
- try {
275
- const externalPath = join(dir, "external.json");
276
- writeFileSync(externalPath, JSON.stringify({
277
- workflowId: "resume_context",
278
- summary: "Managed resume_context recovered 1 stable item(s), 0 pattern(s), and 0 case(s).",
279
- source: "managed",
280
- task: "A2A code Claude SDK calling error",
281
- recordedAt: "2026-03-18T00:01:00.000Z",
282
- observationId: "obs-external",
283
- resolvedScope: "global",
284
- }, null, 2) + "\n");
285
-
286
- const review = reviewWorkflowObservationScopes({
287
- dir,
288
- currentScope: "global",
289
- source: "managed",
290
- workflowId: "resume_context",
291
- });
292
- expect(review.totalCount).toBe(1);
293
- expect(review.manualReviewCount).toBe(1);
294
-
295
- const result = keepWorkflowObservationScopesGlobal(review, {
296
- ids: ["obs-external"],
297
- reason: "external-task",
298
- reviewedAt: "2026-03-18T00:05:00.000Z",
299
- });
300
- expect(result.reviewedCount).toBe(1);
301
- expect(result.skippedCount).toBe(0);
302
- expect(result.unmatchedIds).toEqual([]);
303
- expect(result.ambiguousIds).toEqual([]);
304
-
305
- const rewritten = JSON.parse(readFileSync(externalPath, "utf-8"));
306
- expect(rewritten.legacyScopeReview).toEqual({
307
- decision: "keep",
308
- kind: "global",
309
- reason: "external-task",
310
- reviewedAt: "2026-03-18T00:05:00.000Z",
311
- });
312
-
313
- const nextReview = reviewWorkflowObservationScopes({
314
- dir,
315
- currentScope: "global",
316
- source: "managed",
317
- workflowId: "resume_context",
318
- });
319
- expect(nextReview.totalCount).toBe(0);
320
- } finally {
321
- rmSync(dir, { recursive: true, force: true });
322
- }
323
- });
324
-
325
- it("keeps mixed rewrite and keep-global review flows converged on a clean second pass", async () => {
326
- const root = mkdtempSync(join(tmpdir(), "recallnest-workflow-scope-mixed-"));
327
- try {
328
- const workflowDir = join(root, "workflow-observations");
329
- mkdirSync(workflowDir, { recursive: true });
330
- const recallnestPath = join(workflowDir, "recallnest.json");
331
- const externalPath = join(workflowDir, "external.json");
332
- writeFileSync(recallnestPath, JSON.stringify({
333
- workflowId: "resume_context",
334
- summary: "Managed resume_context recovered 1 stable item(s), 1 pattern(s), and 1 case(s).",
335
- source: "managed",
336
- task: "Continue RecallNest scope cleanup",
337
- recordedAt: "2026-03-18T00:00:00.000Z",
338
- observationId: "obs-recallnest",
339
- resolvedScope: "global",
340
- }, null, 2) + "\n");
341
- writeFileSync(externalPath, JSON.stringify({
342
- workflowId: "resume_context",
343
- summary: "Managed resume_context recovered 1 stable item(s), 0 pattern(s), and 0 case(s).",
344
- source: "managed",
345
- task: "A2A code Claude SDK calling error",
346
- recordedAt: "2026-03-18T00:01:00.000Z",
347
- observationId: "obs-external",
348
- resolvedScope: "global",
349
- }, null, 2) + "\n");
350
-
351
- const review = reviewWorkflowObservationScopes({
352
- dir: workflowDir,
353
- currentScope: "global",
354
- source: "managed",
355
- workflowId: "resume_context",
356
- cues: [parseWorkflowObservationScopeCue("recallnest=project:recallnest")],
357
- });
358
- expect(review.totalCount).toBe(2);
359
- expect(review.suggestedCount).toBe(1);
360
- expect(review.manualReviewCount).toBe(1);
361
-
362
- const keepResult = keepWorkflowObservationScopesGlobal(review, {
363
- ids: ["obs-external"],
364
- reason: "external-task",
365
- reviewedAt: "2026-03-18T00:05:00.000Z",
366
- });
367
- const rewriteResult = applyWorkflowObservationScopeSuggestions(review);
368
- expect(keepResult.reviewedCount).toBe(1);
369
- expect(rewriteResult.updatedCount).toBe(1);
370
-
371
- const nextReview = reviewWorkflowObservationScopes({
372
- dir: workflowDir,
373
- currentScope: "global",
374
- source: "managed",
375
- workflowId: "resume_context",
376
- cues: [parseWorkflowObservationScopeCue("recallnest=project:recallnest")],
377
- });
378
- expect(nextReview.totalCount).toBe(0);
379
-
380
- const store = new MemoryStore({
381
- dbPath: join(root, "db"),
382
- vectorDim: 3,
383
- });
384
- const inventory = await collectScopeInventory({
385
- store,
386
- sampleLimit: 5,
387
- workflowObservationsDir: workflowDir,
388
- pinsDir: join(root, "pins"),
389
- checkpointsDir: join(root, "checkpoints"),
390
- });
391
- const workflowLayer = inventory.layers.find((layer) => layer.layer === "workflow-observations");
392
- expect(inventory.totalAnomalyCount).toBe(0);
393
- expect(inventory.totalReviewedCount).toBe(1);
394
- expect(workflowLayer?.anomalyCount).toBe(0);
395
- expect(workflowLayer?.reviewedCount).toBe(1);
396
- } finally {
397
- rmSync(root, { recursive: true, force: true });
398
- }
399
- });
400
- });