Documentation > Source Code
Source Code Browser
Browse the gsd-skill-creator codebase module by module. Each page shows the module's purpose, file tree, key TypeScript interfaces, and links to related documentation. The codebase contains 51 modules, 1,546 TypeScript files, and 119 directories under src/.
Architecture Layers
Modules are organized into 6 dependency layers. Modules only depend on layers below them.
| Layer | Modules | Purpose |
|---|---|---|
| L1: Foundation | types/, identifiers/, portability/, events/, roles/, capabilities/ | Type definitions, constants, interfaces |
| L2: Infrastructure | storage/, validation/, embeddings/, testing/, safety/, disclosure/ | CRUD, validation, security, testing |
| L3: Analysis | observation/, detection/, composition/, agents/, teams/, activation/, conflicts/, discovery/, brainstorm/, retrieval/, evaluator/, knowledge/ | Pattern detection, agent composition, skill analysis |
| L4: Core Processing | application/, simulation/, learning/, calibration/ | 6-stage pipeline, bounded refinement, threshold optimization |
| L5: Orchestration | workflows/, skill-workflows/, orchestrator/, initialization/, integration/, staging/, vtm/ | Multi-step coordination, GSD integration |
| L6: Entry Points | cli/, hooks/, mcp/, launcher/, terminal/, console/, dashboard/ | User-facing interfaces |
| Specialized | chipset/, den/, amiga/, aminet/, agc/, electronics-pack/, cloud-ops/, bundles/ | Domain-specific systems |
Module Index
| Module | Source Files | Tests | Description |
|---|---|---|---|
activation/ | 5 | 3 | Activation likelihood scoring and improvement suggestions |
agc/ | 36 | 33 | Apollo Guidance Computer educational simulation |
agents/ | 5 | 3 | Agent composition from co-activated skill clusters |
amiga/ | 45 | 42 | AMIGA platform integration and chipset extensions |
aminet/ | 39 | 38 | Aminet-style archive indexing and distribution |
application/ | 14 | 6 | 6-stage skill loading pipeline (Score, Resolve, ModelFilter, CacheOrder, Budget, Load) |
brainstorm/ | 39 | 13 | Multi-agent structured ideation with 16 techniques |
bundles/ | 6 | 5 | Skill bundle packaging and distribution |
calibration/ | 7 | 5 | Threshold optimization using real activation data |
capabilities/ | 14 | 12 | Domain-specific capability declarations |
chipset/ | 33 | 23 | Multi-agent chipset architecture with ISA encoding |
cli/ | 45 | 25 | Command-line interface for all skill-creator operations |
cloud-ops/ | 16 | 7 | Cloud infrastructure management and deployment |
composition/ | 5 | 4 | Skill inheritance, dependency graphs, cycle detection |
conflicts/ | 4 | 3 | Semantic overlap detection between skills |
console/ | 13 | 16 | Filesystem message bus for inter-process communication |
dashboard/ | 79 | 70 | Web dashboard for planning docs visualization |
den/ | 20 | 24 | Agent configuration and staff topology management |
detection/ | 6 | 3 | Pattern analysis, skill generation, suggestion management |
disclosure/ | 6 | 6 | Progressive disclosure for large skills (>2000 words) |
discovery/ | 24 | 21 | DBSCAN clustering with cosine distance, k-NN knee detection |
electronics-pack/ | 32 | 38 | Hardware simulation with chip-level accuracy (5 engines, 77 labs) |
embeddings/ | 5 | 4 | Semantic analysis with 3-tier embedding strategy |
evaluator/ | 5 | 4 | Skill quality evaluation and scoring |
events/ | 6 | 5 | Event system for cross-module communication |
hooks/ | 4 | 2 | Claude Code session lifecycle integration |
identifiers/ | 5 | 3 | Unique identifier generation and validation |
initialization/ | 1 | 0 | Project and environment bootstrap |
integration/ | 13 | 8 | Integration config, install scripts, session wiring |
knowledge/ | 21 | 22 | Knowledge graph structures for domain modeling |
launcher/ | 4 | 2 | System startup and process management |
learning/ | 7 | 6 | Bounded refinement engine with drift tracking |
mcp/ | 59 | 39 | Model Context Protocol gateway, bridge, and security |
observation/ | 18 | 18 | Session observation, transcript parsing, pattern summarization |
orchestrator/ | 53 | 37 | GSD command routing with 5-stage intent classification |
portability/ | 4 | 3 | Cross-platform skill distribution |
retrieval/ | 5 | 3 | Adaptive routing between TF-IDF and embedding search |
roles/ | 5 | 4 | Skill role assignment patterns |
safety/ | 4 | 4 | Defense-in-depth security coordination |
simulation/ | 7 | 3 | Predict skill activation without invoking Claude |
skill-workflows/ | 8 | 7 | Multi-step skill operations |
staging/ | 51 | 48 | Staging pipeline with state machine and intake queue |
storage/ | 4 | 2 | File-based CRUD with SKILL.md format |
teams/ | 9 | 8 | 5 team topologies: leader-worker, pipeline, swarm, router, map-reduce |
terminal/ | 6 | 4 | Terminal UI rendering and Wetty integration |
testing/ | 11 | 3 | Test infrastructure: stores, runners, formatters |
types/ | 21 | 2 | All TypeScript interfaces, types, and constants |
validation/ | 17 | 18 | 17 validators: safety, budget, schema, specialized |
vtm/ | 17 | 17 | Vision-to-mission phase modeling and execution |
workflows/ | 4 | 0 | High-level workflow orchestration |
Project Stats
| Metric | Count |
|---|---|
| Top-level modules | 51 |
| Total directories | 119 |
| TypeScript source files | 1,546 |
| Tests (12,545 total) | Vitest |
| Lines of code | ~410,000 |
| Versions released | 33 (v1.0 – v1.33) |
Quick Links
- Architecture Overview — 6-layer dependency hierarchy
- System Layers — All modules with TypeScript interfaces
- Data Flows — Three core pipelines with real function calls
- Extending the System — Custom validators, storage backends, pipeline stages