agentic-flow 2.0.1-alpha.16 → 2.0.1-alpha.17

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 CHANGED
@@ -2,6 +2,48 @@
2
2
 
3
3
  All notable changes to this project will be documented in this file.
4
4
 
5
+ ## [2.0.1-alpha.17] - 2025-12-31
6
+
7
+ ### Added
8
+ - **10 Attention Mechanisms** (ruvector@0.1.63):
9
+ - `MultiHeadAttention` - Pattern matching
10
+ - `FlashAttention` - Memory-efficient O(n)
11
+ - `HyperbolicAttention` - Hierarchical data (AST, trees)
12
+ - `LinearAttention` - O(n) real-time hooks
13
+ - `LocalGlobalAttention` - Long document summarization
14
+ - `MoEAttention` - Multi-agent routing
15
+ - `GraphRoPeAttention` - Code structure analysis
16
+ - `EdgeFeaturedAttention` - Import/dependency graphs
17
+ - `DualSpaceAttention` - Euclidean + Hyperbolic
18
+
19
+ - **Extended Worker Pool Operations**:
20
+ - `speculativeEmbed(files)` - Pre-embed for post-edit
21
+ - `analyzeAST(files)` - Multi-file parallel AST
22
+ - `analyzeComplexity(files)` - Cyclomatic metrics
23
+ - `buildDependencyGraph(files)` - Import graph
24
+ - `securityScan(files)` - Parallel SAST
25
+ - `ragRetrieve(query, chunks)` - Parallel RAG
26
+ - `rankContext(query, contexts)` - Relevance ranking
27
+ - `deduplicate(texts)` - Semantic deduplication
28
+ - `gitBlame(files)` - Parallel blame analysis
29
+ - `gitChurn(patterns, since)` - Code churn metrics
30
+
31
+ - **Parallel Attention Compute**:
32
+ - `parallelAttentionCompute()` - 3-4x faster multi-query
33
+ - `batchAttentionCompute()` - Batch Q-K-V sets
34
+ - `computeFlashAttentionAsync()` - Non-blocking flash
35
+ - `getAttentionForUseCase()` - Auto-select attention type
36
+
37
+ - **Training Utilities**:
38
+ - `AdamOptimizer` - Attention parameter training
39
+ - `infoNceLoss()` - Contrastive learning
40
+ - `mineHardNegatives()` - Hard negative mining
41
+ - `benchmarkAttention()` - Performance comparison
42
+
43
+ ### Changed
44
+ - Updated ruvector dependency: 0.1.62 → 0.1.63
45
+ - Intelligence bridge supports all attention mechanisms
46
+
5
47
  ## [2.0.1-alpha.16] - 2025-12-31
6
48
 
7
49
  ### Added