datdang-dev/transformation-engine
studio/core/transformation-engine/SKILL.md
Director K's orchestration layer — takes forensic-state.json from any upstream source (manga forensics, Ren'Py extraction), runs knowledge injection, delegates to lewd-writer and quality-audit, manages rewrite loop until score >= 85. Input-agnostic; any source producing valid forensic-state.json can plug in.
development
Updated May 5, 2026
$ install --global
skillsauthnpx skillsauth add datdang-dev/light-novel-development transformation-engineInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 5, 2026, 3:53 AM333.2s4 files scanned
SKILL.md
- name:
- transformation-engine
- description:
- Director K's orchestration layer — takes forensic-state.json from any upstream source (manga forensics, Ren'Py extraction), runs knowledge injection, delegates to lewd-writer and quality-audit, manages rewrite loop until score >= 85. Input-agnostic; any source producing valid forensic-state.json can plug in.
- owner:
- datdang
- version:
- 1.0.0
- tags:
- [orchestration, pipeline, director, rewrite-loop, rag]
- - scene_tag:
- explicit|r18|adaptation|transformation
- - path:
- {{project_root}}/studio/knowledge/glossaries/hentai_lexicon.md
- modules:
- []
Transformation Engine
Overview
The Transformation Engine is the central orchestration layer that bridges forensic analysis and final prose output. Operated by Director K, it takes a standardized forensic-state.json (from any source — Kana's manga forensics or Ren'Py extraction) and produces a verified draft-prose.json through a 4-phase pipeline:
- Knowledge Injection — JIT RAG query based on
content_tags - Prose Generation — Delegates to Lewd Writer (Suki)
- Quality Audit — Delegates to Gooner Editor (Riko)
- Rewrite Loop — Score < 85 triggers rewrite; on PASS, commits continuity state
This engine is input-agnostic by design — any upstream source that produces a valid forensic-state.json can plug into it.
On Activation
- Load
forensic-state.jsonfrom the provided path - Verify all schemas exist in
{{project_root}}/studio/schemas/ - Execute knowledge injection script (step 1)
- Delegate to Lewd Writer engine, then Quality Audit engine
- Manage rewrite loop until PASS
Steps
| Step | File | Purpose |
|------|------|---------|
| 1 | steps/step-01-knowledge-injection.md | Run knowledge_injector.py to generate knowledge_payload.md |
Steps 2–4 are delegated to the Lewd Writer and Quality Audit engines respectively.
Scripts
| File | Purpose |
|------|---------|
| knowledge_injector.py | Python script — queries knowledge base using content_tags from forensic state, outputs knowledge_payload.md |
Usage:
python3 {{project_root}}/studio/core/transformation-engine/knowledge_injector.py \
{forensic_state_path} \
{payload_output_path} \
{{project_root}}/studio/knowledge/
Dependencies
- Orchestrator: Director K (
DIR—lnd-orchestrator.agent.yaml) - Input Schema:
forensic-state.schema.json - Output Schema:
draft-prose.schema.json,audit-report.schema.json - Sub-Engines:
core/lewd-writer,services/quality-audit - Knowledge Base:
{{project_root}}/studio/knowledge/(fetish-db, glossaries, style-guides)
Quick Reference
| Intent | Trigger | Route |
|--------|---------|-------|
| Full transformation | Invoked by Gooner Alchemist pipeline | Execute steps 1 → delegate to Lewd Writer → Quality Audit |
| Knowledge injection only | Manual run | Execute knowledge_injector.py standalone |
Related Skills
datdang-dev/sillytavern-expert
development
VerifiedTrustedCommunity
SillyTavern framework engineering expert — Prompt Manager architecture, character card V3 optimization, context template engineering, lorebook design, and Vietnamese R18 roleplay configuration. Bridges LND Studio character bibles to high-fidelity ST deployments.
SKILL.mdUpdated May 29, 2026
datdang-dev/r18-researcher
documentation
VerifiedTrustedCommunity
Kuro's R18 Research & Downloader Engine. Harvests fetish references, downloads R18 assets (novels, manga, illustrations), and extracts natural Japanese R18 prose style guides from Syosetu.
SKILL.mdUpdated May 25, 2026
datdang-dev/novel-development
development
VerifiedTrustedCommunity
Ultimate Novel Development Pipeline — A slow-paced, deeply contextualized narrative engine. Orchestrated flow: Kana (Visuals) -> Luna (World/Pacing) -> Suki (Long-form Prose).
SKILL.mdUpdated May 5, 2026
datdang-dev/scene-prelude
data-ai
VerifiedTrustedCommunity
Luna's Scene Prelude engine — generates an erotic narrative context (micro-scenario) from Kana's forensic data, bridging raw visual analysis to Suki's caption writing. Outputs a structured scenario seed that maximizes erotic tension and narrative coherence.
SKILL.mdUpdated May 5, 2026