@londer/cortex 0.1.0
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.
- package/CHANGELOG.md +62 -0
- package/LICENSE +21 -0
- package/README.md +183 -0
- package/dist/cli.d.ts +3 -0
- package/dist/cli.d.ts.map +1 -0
- package/dist/cli.js +278 -0
- package/dist/cli.js.map +1 -0
- package/dist/config-store.d.ts +15 -0
- package/dist/config-store.d.ts.map +1 -0
- package/dist/config-store.js +86 -0
- package/dist/config-store.js.map +1 -0
- package/dist/config.d.ts +28 -0
- package/dist/config.d.ts.map +1 -0
- package/dist/config.js +39 -0
- package/dist/config.js.map +1 -0
- package/dist/consolidation/cluster-finder.d.ts +5 -0
- package/dist/consolidation/cluster-finder.d.ts.map +1 -0
- package/dist/consolidation/cluster-finder.js +43 -0
- package/dist/consolidation/cluster-finder.js.map +1 -0
- package/dist/consolidation/consolidator.d.ts +29 -0
- package/dist/consolidation/consolidator.d.ts.map +1 -0
- package/dist/consolidation/consolidator.js +296 -0
- package/dist/consolidation/consolidator.js.map +1 -0
- package/dist/embedding/embedder.d.ts +9 -0
- package/dist/embedding/embedder.d.ts.map +1 -0
- package/dist/embedding/embedder.js +89 -0
- package/dist/embedding/embedder.js.map +1 -0
- package/dist/extraction/auto-extractor.d.ts +7 -0
- package/dist/extraction/auto-extractor.d.ts.map +1 -0
- package/dist/extraction/auto-extractor.js +174 -0
- package/dist/extraction/auto-extractor.js.map +1 -0
- package/dist/extraction/conversation-buffer.d.ts +15 -0
- package/dist/extraction/conversation-buffer.d.ts.map +1 -0
- package/dist/extraction/conversation-buffer.js +60 -0
- package/dist/extraction/conversation-buffer.js.map +1 -0
- package/dist/extraction/extractor.d.ts +23 -0
- package/dist/extraction/extractor.d.ts.map +1 -0
- package/dist/extraction/extractor.js +108 -0
- package/dist/extraction/extractor.js.map +1 -0
- package/dist/extraction/tier1-regex.d.ts +7 -0
- package/dist/extraction/tier1-regex.d.ts.map +1 -0
- package/dist/extraction/tier1-regex.js +119 -0
- package/dist/extraction/tier1-regex.js.map +1 -0
- package/dist/extraction/tier2-nlp.d.ts +13 -0
- package/dist/extraction/tier2-nlp.d.ts.map +1 -0
- package/dist/extraction/tier2-nlp.js +195 -0
- package/dist/extraction/tier2-nlp.js.map +1 -0
- package/dist/extraction/tier3-llm.d.ts +8 -0
- package/dist/extraction/tier3-llm.d.ts.map +1 -0
- package/dist/extraction/tier3-llm.js +57 -0
- package/dist/extraction/tier3-llm.js.map +1 -0
- package/dist/index.d.ts +2 -0
- package/dist/index.d.ts.map +1 -0
- package/dist/index.js +398 -0
- package/dist/index.js.map +1 -0
- package/dist/llm/anthropic-client.d.ts +18 -0
- package/dist/llm/anthropic-client.d.ts.map +1 -0
- package/dist/llm/anthropic-client.js +116 -0
- package/dist/llm/anthropic-client.js.map +1 -0
- package/dist/orchestration/ranker.d.ts +18 -0
- package/dist/orchestration/ranker.d.ts.map +1 -0
- package/dist/orchestration/ranker.js +124 -0
- package/dist/orchestration/ranker.js.map +1 -0
- package/dist/orchestration/scope.d.ts +15 -0
- package/dist/orchestration/scope.d.ts.map +1 -0
- package/dist/orchestration/scope.js +28 -0
- package/dist/orchestration/scope.js.map +1 -0
- package/dist/storage/neo4j.d.ts +19 -0
- package/dist/storage/neo4j.d.ts.map +1 -0
- package/dist/storage/neo4j.js +246 -0
- package/dist/storage/neo4j.js.map +1 -0
- package/dist/storage/qdrant.d.ts +21 -0
- package/dist/storage/qdrant.d.ts.map +1 -0
- package/dist/storage/qdrant.js +125 -0
- package/dist/storage/qdrant.js.map +1 -0
- package/dist/storage/sqlite.d.ts +23 -0
- package/dist/storage/sqlite.d.ts.map +1 -0
- package/dist/storage/sqlite.js +162 -0
- package/dist/storage/sqlite.js.map +1 -0
- package/dist/templates/claude-instructions.d.ts +11 -0
- package/dist/templates/claude-instructions.d.ts.map +1 -0
- package/dist/templates/claude-instructions.js +110 -0
- package/dist/templates/claude-instructions.js.map +1 -0
- package/dist/tools/config.d.ts +21 -0
- package/dist/tools/config.d.ts.map +1 -0
- package/dist/tools/config.js +96 -0
- package/dist/tools/config.js.map +1 -0
- package/dist/tools/consolidate.d.ts +4 -0
- package/dist/tools/consolidate.d.ts.map +1 -0
- package/dist/tools/consolidate.js +16 -0
- package/dist/tools/consolidate.js.map +1 -0
- package/dist/tools/context.d.ts +8 -0
- package/dist/tools/context.d.ts.map +1 -0
- package/dist/tools/context.js +84 -0
- package/dist/tools/context.js.map +1 -0
- package/dist/tools/forget.d.ts +7 -0
- package/dist/tools/forget.d.ts.map +1 -0
- package/dist/tools/forget.js +49 -0
- package/dist/tools/forget.js.map +1 -0
- package/dist/tools/graph-query.d.ts +5 -0
- package/dist/tools/graph-query.d.ts.map +1 -0
- package/dist/tools/graph-query.js +35 -0
- package/dist/tools/graph-query.js.map +1 -0
- package/dist/tools/ingest.d.ts +8 -0
- package/dist/tools/ingest.d.ts.map +1 -0
- package/dist/tools/ingest.js +101 -0
- package/dist/tools/ingest.js.map +1 -0
- package/dist/tools/relate.d.ts +4 -0
- package/dist/tools/relate.d.ts.map +1 -0
- package/dist/tools/relate.js +10 -0
- package/dist/tools/relate.js.map +1 -0
- package/dist/tools/search.d.ts +6 -0
- package/dist/tools/search.d.ts.map +1 -0
- package/dist/tools/search.js +43 -0
- package/dist/tools/search.js.map +1 -0
- package/dist/tools/store.d.ts +8 -0
- package/dist/tools/store.d.ts.map +1 -0
- package/dist/tools/store.js +64 -0
- package/dist/tools/store.js.map +1 -0
- package/dist/types/index.d.ts +289 -0
- package/dist/types/index.d.ts.map +1 -0
- package/dist/types/index.js +3 -0
- package/dist/types/index.js.map +1 -0
- package/docker-compose.yml +27 -0
- package/package.json +71 -0
|
@@ -0,0 +1,195 @@
|
|
|
1
|
+
import nlp from 'compromise';
|
|
2
|
+
const SIMILARITY_DEDUP_THRESHOLD = 0.85;
|
|
3
|
+
// Technical terms that signal entity relevance
|
|
4
|
+
const TECH_PATTERNS = [
|
|
5
|
+
/(?:api|sdk|cli|gui|ui|ux)\b/i,
|
|
6
|
+
/(?:service|server|client|module|component|middleware|handler|controller)\b/i,
|
|
7
|
+
/(?:database|cache|queue|store|repository|factory|adapter)\b/i,
|
|
8
|
+
/(?:auth|oauth|jwt|token|session)\b/i,
|
|
9
|
+
/(?:docker|kubernetes|nginx|redis|postgres|mongo)\b/i,
|
|
10
|
+
];
|
|
11
|
+
export class Tier2NLP {
|
|
12
|
+
embedder;
|
|
13
|
+
qdrant;
|
|
14
|
+
constructor(embedder = null, qdrant = null) {
|
|
15
|
+
this.embedder = embedder;
|
|
16
|
+
this.qdrant = qdrant;
|
|
17
|
+
}
|
|
18
|
+
async extract(text, _project) {
|
|
19
|
+
const doc = nlp(text);
|
|
20
|
+
const entities = this.extractEntities(doc, text);
|
|
21
|
+
const relationships = this.extractRelationships(doc, text);
|
|
22
|
+
// Dedup entities against existing ones in Qdrant via embedding similarity
|
|
23
|
+
const dedupedEntities = await this.dedupEntities(entities);
|
|
24
|
+
return {
|
|
25
|
+
entities: dedupedEntities,
|
|
26
|
+
relationships,
|
|
27
|
+
tier: 2,
|
|
28
|
+
};
|
|
29
|
+
}
|
|
30
|
+
extractEntities(doc, text) {
|
|
31
|
+
const seen = new Set();
|
|
32
|
+
const entities = [];
|
|
33
|
+
// Extract noun phrases
|
|
34
|
+
const nouns = doc.nouns().out('array');
|
|
35
|
+
for (const noun of nouns) {
|
|
36
|
+
const name = toSnakeCase(noun);
|
|
37
|
+
if (seen.has(name) || name.length < 3 || isStopPhrase(noun))
|
|
38
|
+
continue;
|
|
39
|
+
seen.add(name);
|
|
40
|
+
entities.push({
|
|
41
|
+
name,
|
|
42
|
+
original: noun,
|
|
43
|
+
type: inferType(noun),
|
|
44
|
+
confidence: 0.6,
|
|
45
|
+
});
|
|
46
|
+
}
|
|
47
|
+
// Extract named entities (people, places, organizations)
|
|
48
|
+
const people = doc.people().out('array');
|
|
49
|
+
for (const person of people) {
|
|
50
|
+
const name = toSnakeCase(person);
|
|
51
|
+
if (seen.has(name) || name.length < 3)
|
|
52
|
+
continue;
|
|
53
|
+
seen.add(name);
|
|
54
|
+
entities.push({
|
|
55
|
+
name,
|
|
56
|
+
original: person,
|
|
57
|
+
type: 'person',
|
|
58
|
+
confidence: 0.6,
|
|
59
|
+
});
|
|
60
|
+
}
|
|
61
|
+
const orgs = doc.organizations().out('array');
|
|
62
|
+
for (const org of orgs) {
|
|
63
|
+
const name = toSnakeCase(org);
|
|
64
|
+
if (seen.has(name) || name.length < 3)
|
|
65
|
+
continue;
|
|
66
|
+
seen.add(name);
|
|
67
|
+
entities.push({
|
|
68
|
+
name,
|
|
69
|
+
original: org,
|
|
70
|
+
type: 'service',
|
|
71
|
+
confidence: 0.6,
|
|
72
|
+
});
|
|
73
|
+
}
|
|
74
|
+
// Also extract code-style identifiers that NLP misses
|
|
75
|
+
const codePatterns = [
|
|
76
|
+
/[a-z][a-z0-9]*(?:_[a-z0-9]+)+/g,
|
|
77
|
+
/[a-z]+(?:[A-Z][a-z0-9]+)+/g,
|
|
78
|
+
/[A-Z][a-z0-9]+(?:[A-Z][a-z0-9]+)+/g,
|
|
79
|
+
];
|
|
80
|
+
for (const pattern of codePatterns) {
|
|
81
|
+
const regex = new RegExp(pattern.source, pattern.flags);
|
|
82
|
+
let match;
|
|
83
|
+
while ((match = regex.exec(text)) !== null) {
|
|
84
|
+
const original = match[0];
|
|
85
|
+
const name = toSnakeCase(original);
|
|
86
|
+
if (seen.has(name) || name.length < 3)
|
|
87
|
+
continue;
|
|
88
|
+
seen.add(name);
|
|
89
|
+
entities.push({
|
|
90
|
+
name,
|
|
91
|
+
original,
|
|
92
|
+
type: inferType(original),
|
|
93
|
+
confidence: 0.5,
|
|
94
|
+
});
|
|
95
|
+
}
|
|
96
|
+
}
|
|
97
|
+
// Boost confidence for entities matching technical patterns
|
|
98
|
+
for (const entity of entities) {
|
|
99
|
+
if (TECH_PATTERNS.some(p => p.test(entity.original))) {
|
|
100
|
+
entity.confidence = Math.min(entity.confidence + 0.1, 1.0);
|
|
101
|
+
}
|
|
102
|
+
}
|
|
103
|
+
return entities;
|
|
104
|
+
}
|
|
105
|
+
extractRelationships(doc, text) {
|
|
106
|
+
const relationships = [];
|
|
107
|
+
const seen = new Set();
|
|
108
|
+
// Verb pattern matching for relationships
|
|
109
|
+
const verbPatterns = [
|
|
110
|
+
{ pattern: /(\b\w[\w\s]{0,30}?\w)\s+depends?\s+on\s+(\b\w[\w\s]{0,30}?\w)/gi, relation: 'DEPENDS_ON' },
|
|
111
|
+
{ pattern: /(\b\w[\w\s]{0,30}?\w)\s+uses?\s+(\b\w[\w\s]{0,30}?\w)/gi, relation: 'USES' },
|
|
112
|
+
{ pattern: /(\b\w[\w\s]{0,30}?\w)\s+implements?\s+(\b\w[\w\s]{0,30}?\w)/gi, relation: 'IMPLEMENTS' },
|
|
113
|
+
{ pattern: /(\b\w[\w\s]{0,30}?\w)\s+calls?\s+(\b\w[\w\s]{0,30}?\w)/gi, relation: 'CALLS' },
|
|
114
|
+
{ pattern: /(\b\w[\w\s]{0,30}?\w)\s+is\s+part\s+of\s+(\b\w[\w\s]{0,30}?\w)/gi, relation: 'PART_OF' },
|
|
115
|
+
{ pattern: /(\b\w[\w\s]{0,30}?\w)\s+connects?\s+to\s+(\b\w[\w\s]{0,30}?\w)/gi, relation: 'RELATED_TO' },
|
|
116
|
+
];
|
|
117
|
+
for (const { pattern, relation } of verbPatterns) {
|
|
118
|
+
const regex = new RegExp(pattern.source, pattern.flags);
|
|
119
|
+
let match;
|
|
120
|
+
while ((match = regex.exec(text)) !== null) {
|
|
121
|
+
const from = toSnakeCase(match[1].trim());
|
|
122
|
+
const to = toSnakeCase(match[2].trim());
|
|
123
|
+
if (from.length < 2 || to.length < 2)
|
|
124
|
+
continue;
|
|
125
|
+
const key = `${from}:${relation}:${to}`;
|
|
126
|
+
if (seen.has(key))
|
|
127
|
+
continue;
|
|
128
|
+
seen.add(key);
|
|
129
|
+
relationships.push({ from, relation, to, confidence: 0.5 });
|
|
130
|
+
}
|
|
131
|
+
}
|
|
132
|
+
return relationships;
|
|
133
|
+
}
|
|
134
|
+
async dedupEntities(entities) {
|
|
135
|
+
if (!this.embedder || !this.qdrant || entities.length === 0) {
|
|
136
|
+
return entities;
|
|
137
|
+
}
|
|
138
|
+
try {
|
|
139
|
+
const names = entities.map(e => e.name);
|
|
140
|
+
const vectors = await this.embedder.embedBatch(names);
|
|
141
|
+
const result = [];
|
|
142
|
+
for (let i = 0; i < entities.length; i++) {
|
|
143
|
+
const similar = await this.qdrant.search(vectors[i], undefined, 1);
|
|
144
|
+
if (similar.length > 0 &&
|
|
145
|
+
similar[0].score > SIMILARITY_DEDUP_THRESHOLD &&
|
|
146
|
+
similar[0].entities.length > 0) {
|
|
147
|
+
// Found a similar existing entity — reuse its name
|
|
148
|
+
const existingName = similar[0].entities[0];
|
|
149
|
+
if (existingName && existingName !== entities[i].name) {
|
|
150
|
+
result.push({
|
|
151
|
+
...entities[i],
|
|
152
|
+
name: existingName,
|
|
153
|
+
});
|
|
154
|
+
continue;
|
|
155
|
+
}
|
|
156
|
+
}
|
|
157
|
+
result.push(entities[i]);
|
|
158
|
+
}
|
|
159
|
+
return result;
|
|
160
|
+
}
|
|
161
|
+
catch {
|
|
162
|
+
// If dedup fails, return entities as-is
|
|
163
|
+
return entities;
|
|
164
|
+
}
|
|
165
|
+
}
|
|
166
|
+
}
|
|
167
|
+
function toSnakeCase(str) {
|
|
168
|
+
return str
|
|
169
|
+
.replace(/([a-z0-9])([A-Z])/g, '$1_$2')
|
|
170
|
+
.replace(/[\s-]+/g, '_')
|
|
171
|
+
.toLowerCase()
|
|
172
|
+
.replace(/_+/g, '_')
|
|
173
|
+
.replace(/^_|_$/g, '');
|
|
174
|
+
}
|
|
175
|
+
function inferType(name) {
|
|
176
|
+
const lower = name.toLowerCase();
|
|
177
|
+
if (lower.includes('service') || lower.includes('server') || lower.includes('api'))
|
|
178
|
+
return 'service';
|
|
179
|
+
if (lower.includes('component') || lower.includes('module') || lower.includes('widget'))
|
|
180
|
+
return 'component';
|
|
181
|
+
if (lower.includes('lib') || lower.includes('package') || lower.includes('sdk'))
|
|
182
|
+
return 'library';
|
|
183
|
+
if (lower.includes('project') || lower.includes('repo'))
|
|
184
|
+
return 'project';
|
|
185
|
+
return 'concept';
|
|
186
|
+
}
|
|
187
|
+
const STOP_PHRASES = new Set([
|
|
188
|
+
'the', 'this', 'that', 'it', 'they', 'we', 'i', 'you', 'he', 'she',
|
|
189
|
+
'something', 'nothing', 'everything', 'anything',
|
|
190
|
+
'way', 'thing', 'stuff', 'lot', 'bit', 'time',
|
|
191
|
+
]);
|
|
192
|
+
function isStopPhrase(str) {
|
|
193
|
+
return STOP_PHRASES.has(str.toLowerCase().trim());
|
|
194
|
+
}
|
|
195
|
+
//# sourceMappingURL=tier2-nlp.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"tier2-nlp.js","sourceRoot":"","sources":["../../src/extraction/tier2-nlp.ts"],"names":[],"mappings":"AAAA,OAAO,GAAG,MAAM,YAAY,CAAC;AAK7B,MAAM,0BAA0B,GAAG,IAAI,CAAC;AAExC,+CAA+C;AAC/C,MAAM,aAAa,GAAG;IACpB,8BAA8B;IAC9B,6EAA6E;IAC7E,8DAA8D;IAC9D,qCAAqC;IACrC,qDAAqD;CACtD,CAAC;AAEF,MAAM,OAAO,QAAQ;IACX,QAAQ,CAAkB;IAC1B,MAAM,CAAqB;IAEnC,YAAY,WAA4B,IAAI,EAAE,SAA6B,IAAI;QAC7E,IAAI,CAAC,QAAQ,GAAG,QAAQ,CAAC;QACzB,IAAI,CAAC,MAAM,GAAG,MAAM,CAAC;IACvB,CAAC;IAED,KAAK,CAAC,OAAO,CAAC,IAAY,EAAE,QAAiB;QAC3C,MAAM,GAAG,GAAG,GAAG,CAAC,IAAI,CAAC,CAAC;QAEtB,MAAM,QAAQ,GAAG,IAAI,CAAC,eAAe,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC;QACjD,MAAM,aAAa,GAAG,IAAI,CAAC,oBAAoB,CAAC,GAAG,EAAE,IAAI,CAAC,CAAC;QAE3D,0EAA0E;QAC1E,MAAM,eAAe,GAAG,MAAM,IAAI,CAAC,aAAa,CAAC,QAAQ,CAAC,CAAC;QAE3D,OAAO;YACL,QAAQ,EAAE,eAAe;YACzB,aAAa;YACb,IAAI,EAAE,CAAC;SACR,CAAC;IACJ,CAAC;IAEO,eAAe,CAAC,GAA2B,EAAE,IAAY;QAC/D,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;QAC/B,MAAM,QAAQ,GAAsB,EAAE,CAAC;QAEvC,uBAAuB;QACvB,MAAM,KAAK,GAAG,GAAG,CAAC,KAAK,EAAE,CAAC,GAAG,CAAC,OAAO,CAAa,CAAC;QACnD,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,MAAM,IAAI,GAAG,WAAW,CAAC,IAAI,CAAC,CAAC;YAC/B,IAAI,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC,IAAI,YAAY,CAAC,IAAI,CAAC;gBAAE,SAAS;YACtE,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;YACf,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI;gBACJ,QAAQ,EAAE,IAAI;gBACd,IAAI,EAAE,SAAS,CAAC,IAAI,CAAC;gBACrB,UAAU,EAAE,GAAG;aAChB,CAAC,CAAC;QACL,CAAC;QAED,yDAAyD;QACzD,MAAM,MAAM,GAAG,GAAG,CAAC,MAAM,EAAE,CAAC,GAAG,CAAC,OAAO,CAAa,CAAC;QACrD,KAAK,MAAM,MAAM,IAAI,MAAM,EAAE,CAAC;YAC5B,MAAM,IAAI,GAAG,WAAW,CAAC,MAAM,CAAC,CAAC;YACjC,IAAI,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC;gBAAE,SAAS;YAChD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;YACf,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI;gBACJ,QAAQ,EAAE,MAAM;gBAChB,IAAI,EAAE,QAAQ;gBACd,UAAU,EAAE,GAAG;aAChB,CAAC,CAAC;QACL,CAAC;QAED,MAAM,IAAI,GAAG,GAAG,CAAC,aAAa,EAAE,CAAC,GAAG,CAAC,OAAO,CAAa,CAAC;QAC1D,KAAK,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;YACvB,MAAM,IAAI,GAAG,WAAW,CAAC,GAAG,CAAC,CAAC;YAC9B,IAAI,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC;gBAAE,SAAS;YAChD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;YACf,QAAQ,CAAC,IAAI,CAAC;gBACZ,IAAI;gBACJ,QAAQ,EAAE,GAAG;gBACb,IAAI,EAAE,SAAS;gBACf,UAAU,EAAE,GAAG;aAChB,CAAC,CAAC;QACL,CAAC;QAED,sDAAsD;QACtD,MAAM,YAAY,GAAG;YACnB,gCAAgC;YAChC,4BAA4B;YAC5B,oCAAoC;SACrC,CAAC;QAEF,KAAK,MAAM,OAAO,IAAI,YAAY,EAAE,CAAC;YACnC,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC;YACxD,IAAI,KAA6B,CAAC;YAClC,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBAC3C,MAAM,QAAQ,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;gBAC1B,MAAM,IAAI,GAAG,WAAW,CAAC,QAAQ,CAAC,CAAC;gBACnC,IAAI,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC;oBAAE,SAAS;gBAChD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;gBACf,QAAQ,CAAC,IAAI,CAAC;oBACZ,IAAI;oBACJ,QAAQ;oBACR,IAAI,EAAE,SAAS,CAAC,QAAQ,CAAC;oBACzB,UAAU,EAAE,GAAG;iBAChB,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,4DAA4D;QAC5D,KAAK,MAAM,MAAM,IAAI,QAAQ,EAAE,CAAC;YAC9B,IAAI,aAAa,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC;gBACrD,MAAM,CAAC,UAAU,GAAG,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC,UAAU,GAAG,GAAG,EAAE,GAAG,CAAC,CAAC;YAC7D,CAAC;QACH,CAAC;QAED,OAAO,QAAQ,CAAC;IAClB,CAAC;IAEO,oBAAoB,CAAC,GAA2B,EAAE,IAAY;QACpE,MAAM,aAAa,GAA4B,EAAE,CAAC;QAClD,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;QAE/B,0CAA0C;QAC1C,MAAM,YAAY,GAAiD;YACjE,EAAE,OAAO,EAAE,iEAAiE,EAAE,QAAQ,EAAE,YAAY,EAAE;YACtG,EAAE,OAAO,EAAE,yDAAyD,EAAE,QAAQ,EAAE,MAAM,EAAE;YACxF,EAAE,OAAO,EAAE,+DAA+D,EAAE,QAAQ,EAAE,YAAY,EAAE;YACpG,EAAE,OAAO,EAAE,0DAA0D,EAAE,QAAQ,EAAE,OAAO,EAAE;YAC1F,EAAE,OAAO,EAAE,kEAAkE,EAAE,QAAQ,EAAE,SAAS,EAAE;YACpG,EAAE,OAAO,EAAE,kEAAkE,EAAE,QAAQ,EAAE,YAAY,EAAE;SACxG,CAAC;QAEF,KAAK,MAAM,EAAE,OAAO,EAAE,QAAQ,EAAE,IAAI,YAAY,EAAE,CAAC;YACjD,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC;YACxD,IAAI,KAA6B,CAAC;YAClC,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBAC3C,MAAM,IAAI,GAAG,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;gBAC1C,MAAM,EAAE,GAAG,WAAW,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;gBACxC,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC;oBAAE,SAAS;gBAC/C,MAAM,GAAG,GAAG,GAAG,IAAI,IAAI,QAAQ,IAAI,EAAE,EAAE,CAAC;gBACxC,IAAI,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;oBAAE,SAAS;gBAC5B,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;gBACd,aAAa,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,QAAQ,EAAE,EAAE,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC,CAAC;YAC9D,CAAC;QACH,CAAC;QAED,OAAO,aAAa,CAAC;IACvB,CAAC;IAEO,KAAK,CAAC,aAAa,CAAC,QAA2B;QACrD,IAAI,CAAC,IAAI,CAAC,QAAQ,IAAI,CAAC,IAAI,CAAC,MAAM,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YAC5D,OAAO,QAAQ,CAAC;QAClB,CAAC;QAED,IAAI,CAAC;YACH,MAAM,KAAK,GAAG,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;YACxC,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,QAAQ,CAAC,UAAU,CAAC,KAAK,CAAC,CAAC;YAEtD,MAAM,MAAM,GAAsB,EAAE,CAAC;YACrC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;gBACzC,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,MAAM,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,SAAS,EAAE,CAAC,CAAC,CAAC;gBACnE,IACE,OAAO,CAAC,MAAM,GAAG,CAAC;oBAClB,OAAO,CAAC,CAAC,CAAC,CAAC,KAAK,GAAG,0BAA0B;oBAC7C,OAAO,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,MAAM,GAAG,CAAC,EAC9B,CAAC;oBACD,mDAAmD;oBACnD,MAAM,YAAY,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC;oBAC5C,IAAI,YAAY,IAAI,YAAY,KAAK,QAAQ,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;wBACtD,MAAM,CAAC,IAAI,CAAC;4BACV,GAAG,QAAQ,CAAC,CAAC,CAAC;4BACd,IAAI,EAAE,YAAY;yBACnB,CAAC,CAAC;wBACH,SAAS;oBACX,CAAC;gBACH,CAAC;gBACD,MAAM,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC;YAC3B,CAAC;YACD,OAAO,MAAM,CAAC;QAChB,CAAC;QAAC,MAAM,CAAC;YACP,wCAAwC;YACxC,OAAO,QAAQ,CAAC;QAClB,CAAC;IACH,CAAC;CACF;AAED,SAAS,WAAW,CAAC,GAAW;IAC9B,OAAO,GAAG;SACP,OAAO,CAAC,oBAAoB,EAAE,OAAO,CAAC;SACtC,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,WAAW,EAAE;SACb,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC;SACnB,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAC;AAC3B,CAAC;AAED,SAAS,SAAS,CAAC,IAAY;IAC7B,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC;IACjC,IAAI,KAAK,CAAC,QAAQ,CAAC,SAAS,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,KAAK,CAAC;QAAE,OAAO,SAAS,CAAC;IACrG,IAAI,KAAK,CAAC,QAAQ,CAAC,WAAW,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,QAAQ,CAAC;QAAE,OAAO,WAAW,CAAC;IAC5G,IAAI,KAAK,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,SAAS,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,KAAK,CAAC;QAAE,OAAO,SAAS,CAAC;IAClG,IAAI,KAAK,CAAC,QAAQ,CAAC,SAAS,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,MAAM,CAAC;QAAE,OAAO,SAAS,CAAC;IAC1E,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,MAAM,YAAY,GAAG,IAAI,GAAG,CAAC;IAC3B,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,GAAG,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK;IAClE,WAAW,EAAE,SAAS,EAAE,YAAY,EAAE,UAAU;IAChD,KAAK,EAAE,OAAO,EAAE,OAAO,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM;CAC9C,CAAC,CAAC;AAEH,SAAS,YAAY,CAAC,GAAW;IAC/B,OAAO,YAAY,CAAC,GAAG,CAAC,GAAG,CAAC,WAAW,EAAE,CAAC,IAAI,EAAE,CAAC,CAAC;AACpD,CAAC"}
|
|
@@ -0,0 +1,8 @@
|
|
|
1
|
+
import type { ExtractionResult, LLMClient } from '../types/index.js';
|
|
2
|
+
export declare class Tier3LLM {
|
|
3
|
+
private llmClient;
|
|
4
|
+
constructor(llmClient: LLMClient);
|
|
5
|
+
isAvailable(): boolean;
|
|
6
|
+
extract(text: string, project?: string): Promise<ExtractionResult | null>;
|
|
7
|
+
}
|
|
8
|
+
//# sourceMappingURL=tier3-llm.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"tier3-llm.d.ts","sourceRoot":"","sources":["../../src/extraction/tier3-llm.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,gBAAgB,EAAE,SAAS,EAAE,MAAM,mBAAmB,CAAC;AAiCrE,qBAAa,QAAQ;IACnB,OAAO,CAAC,SAAS,CAAY;gBAEjB,SAAS,EAAE,SAAS;IAIhC,WAAW,IAAI,OAAO;IAIhB,OAAO,CAAC,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC,gBAAgB,GAAG,IAAI,CAAC;CAwChF"}
|
|
@@ -0,0 +1,57 @@
|
|
|
1
|
+
const SYSTEM_PROMPT = `You are an entity extraction system for a developer knowledge base.
|
|
2
|
+
|
|
3
|
+
Extract entities and relationships from the following text. Return JSON only.
|
|
4
|
+
|
|
5
|
+
Rules:
|
|
6
|
+
- Entity names must be snake_case (e.g., "the auth service" → "auth_service")
|
|
7
|
+
- Only extract meaningful technical entities, not generic words
|
|
8
|
+
- Relationships must use: DEPENDS_ON, USES, PART_OF, RELATED_TO, IMPLEMENTS, CALLS
|
|
9
|
+
- Confidence: 1.0 = explicitly stated, 0.7 = strongly implied, 0.4 = weakly implied
|
|
10
|
+
|
|
11
|
+
Respond with only valid JSON:
|
|
12
|
+
{
|
|
13
|
+
"entities": [{"name": "string", "original": "string", "type": "component|service|library|concept|person|project|other", "confidence": number}],
|
|
14
|
+
"relationships": [{"from": "string", "relation": "string", "to": "string", "confidence": number}]
|
|
15
|
+
}`;
|
|
16
|
+
export class Tier3LLM {
|
|
17
|
+
llmClient;
|
|
18
|
+
constructor(llmClient) {
|
|
19
|
+
this.llmClient = llmClient;
|
|
20
|
+
}
|
|
21
|
+
isAvailable() {
|
|
22
|
+
return this.llmClient.isAvailable();
|
|
23
|
+
}
|
|
24
|
+
async extract(text, project) {
|
|
25
|
+
if (!this.isAvailable()) {
|
|
26
|
+
return null;
|
|
27
|
+
}
|
|
28
|
+
const contextLine = project ? `\nProject context: ${project}\n` : '';
|
|
29
|
+
const userContent = `${contextLine}\nText: ${text}`;
|
|
30
|
+
const response = await this.llmClient.extractJSON(SYSTEM_PROMPT, userContent);
|
|
31
|
+
if (!response || !response.entities) {
|
|
32
|
+
return null;
|
|
33
|
+
}
|
|
34
|
+
const validTypes = new Set(['component', 'service', 'library', 'concept', 'person', 'project', 'other']);
|
|
35
|
+
const validRelations = new Set(['DEPENDS_ON', 'USES', 'PART_OF', 'RELATED_TO', 'IMPLEMENTS', 'CALLS']);
|
|
36
|
+
return {
|
|
37
|
+
entities: (response.entities || [])
|
|
38
|
+
.filter(e => e.name && e.original && typeof e.confidence === 'number')
|
|
39
|
+
.map(e => ({
|
|
40
|
+
name: e.name,
|
|
41
|
+
original: e.original,
|
|
42
|
+
type: validTypes.has(e.type) ? e.type : 'other',
|
|
43
|
+
confidence: Math.max(0, Math.min(1, e.confidence)),
|
|
44
|
+
})),
|
|
45
|
+
relationships: (response.relationships || [])
|
|
46
|
+
.filter(r => r.from && r.to && validRelations.has(r.relation) && typeof r.confidence === 'number')
|
|
47
|
+
.map(r => ({
|
|
48
|
+
from: r.from,
|
|
49
|
+
relation: r.relation,
|
|
50
|
+
to: r.to,
|
|
51
|
+
confidence: Math.max(0, Math.min(1, r.confidence)),
|
|
52
|
+
})),
|
|
53
|
+
tier: 3,
|
|
54
|
+
};
|
|
55
|
+
}
|
|
56
|
+
}
|
|
57
|
+
//# sourceMappingURL=tier3-llm.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"tier3-llm.js","sourceRoot":"","sources":["../../src/extraction/tier3-llm.ts"],"names":[],"mappings":"AAEA,MAAM,aAAa,GAAG;;;;;;;;;;;;;;EAcpB,CAAC;AAiBH,MAAM,OAAO,QAAQ;IACX,SAAS,CAAY;IAE7B,YAAY,SAAoB;QAC9B,IAAI,CAAC,SAAS,GAAG,SAAS,CAAC;IAC7B,CAAC;IAED,WAAW;QACT,OAAO,IAAI,CAAC,SAAS,CAAC,WAAW,EAAE,CAAC;IACtC,CAAC;IAED,KAAK,CAAC,OAAO,CAAC,IAAY,EAAE,OAAgB;QAC1C,IAAI,CAAC,IAAI,CAAC,WAAW,EAAE,EAAE,CAAC;YACxB,OAAO,IAAI,CAAC;QACd,CAAC;QAED,MAAM,WAAW,GAAG,OAAO,CAAC,CAAC,CAAC,sBAAsB,OAAO,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC;QACrE,MAAM,WAAW,GAAG,GAAG,WAAW,WAAW,IAAI,EAAE,CAAC;QAEpD,MAAM,QAAQ,GAAG,MAAM,IAAI,CAAC,SAAS,CAAC,WAAW,CAC/C,aAAa,EACb,WAAW,CACZ,CAAC;QAEF,IAAI,CAAC,QAAQ,IAAI,CAAC,QAAQ,CAAC,QAAQ,EAAE,CAAC;YACpC,OAAO,IAAI,CAAC;QACd,CAAC;QAED,MAAM,UAAU,GAAG,IAAI,GAAG,CAAC,CAAC,WAAW,EAAE,SAAS,EAAE,SAAS,EAAE,SAAS,EAAE,QAAQ,EAAE,SAAS,EAAE,OAAO,CAAC,CAAC,CAAC;QACzG,MAAM,cAAc,GAAG,IAAI,GAAG,CAAC,CAAC,YAAY,EAAE,MAAM,EAAE,SAAS,EAAE,YAAY,EAAE,YAAY,EAAE,OAAO,CAAC,CAAC,CAAC;QAEvG,OAAO;YACL,QAAQ,EAAE,CAAC,QAAQ,CAAC,QAAQ,IAAI,EAAE,CAAC;iBAChC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,QAAQ,IAAI,OAAO,CAAC,CAAC,UAAU,KAAK,QAAQ,CAAC;iBACrE,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;gBACT,IAAI,EAAE,CAAC,CAAC,IAAI;gBACZ,QAAQ,EAAE,CAAC,CAAC,QAAQ;gBACpB,IAAI,EAAE,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAA+C,CAAC,CAAC,CAAC,OAAO;gBAC1F,UAAU,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,UAAU,CAAC,CAAC;aACnD,CAAC,CAAC;YACL,aAAa,EAAE,CAAC,QAAQ,CAAC,aAAa,IAAI,EAAE,CAAC;iBAC1C,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,EAAE,IAAI,cAAc,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,CAAC,IAAI,OAAO,CAAC,CAAC,UAAU,KAAK,QAAQ,CAAC;iBACjG,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;gBACT,IAAI,EAAE,CAAC,CAAC,IAAI;gBACZ,QAAQ,EAAE,CAAC,CAAC,QAAQ;gBACpB,EAAE,EAAE,CAAC,CAAC,EAAE;gBACR,UAAU,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,UAAU,CAAC,CAAC;aACnD,CAAC,CAAC;YACL,IAAI,EAAE,CAAC;SACR,CAAC;IACJ,CAAC;CACF"}
|
package/dist/index.d.ts
ADDED
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":""}
|