DBCatalog | Goals & Features Roadmap

Ecosystem High-Level Goals

The Deckard Boundary: Strictly isolate subjective AI writing from hard, deterministic Python state management. Stop the AI from hallucinating databases.
Academic Rigor vs. Playability: Bridge the gap between the "Cathedral" (rigid scholarly reading environments) and the "Laboratory" (playable, interactive digital humanities tools).
No Wheels Reinvented: Continuously adapt proven methodologies from leading external DH projects (e.g., Linked Open Data, Diplomatic transcriptions, statistical network graphs).
Zero-Token Linting: Automatically scrub "AI slop" from our generated writing to maintain a highly dense, critical-reportorial scholarly voice.

Procedural Engine

An interactive Digital Humanities laboratory engine focused on Sir Kenelm Digby, alchemy, and 17th-century trade.

Allow players to experience Renaissance material science by interactively running alchemical recipes.
Ground all gameplay mechanics strictly in academic texts housed within `RenMagDB`.
Develop a playable UI that doesn't feel like a standard academic portal.

Game states currently risk amnesia if the AI isn't forced to write back to the deterministic Python scripts.
The user interface is too heavily reliant on text boxes rather than interactive spatial elements.

As seen in *The Chymistry of Isaac Newton*, recreating recipes physically/digitally is a valid form of academic scholarship, not just a game.
The "Earnestness through Irony" approach makes dense history accessible.

Wire the `locations.json` to Linked Open Data (LOD) URIs for spatial mapping.
Implement "Citational Tethers" to explicitly surface the academic source of an item when a player hovers over it in their inventory.

Knowledge Portal

The definitive scholarly reading environment for the works and theology of Philip K. Dick.

"Lower heaven into reach" by making PKD's massively complex *Exegesis* searchable and structured.
Map out the arguments between different literary scholars regarding PKD's work.

React/HTML static frontend.
Rigid "Scholar Profiles" mapping intellectual lineage and specific literary arguments.
Relational database connecting chapters to specific theological concepts.

The Over-Engineering Trap: The schema is too relational, forcing the AI to hallucinate filler text to satisfy the database joins.
Artificially separates his sci-fi fiction from his lived theological paranoia (2-3-74).

Academic writing must *never flatten contradictions*. If two scholars disagree, the portal must explicitly highlight the fight rather than synthesizing a middle ground.

Add a `thematic_bleed` entity linking the 2-3-74 visions directly back to his earlier fiction.
Run the zero-token `lint_writing.py` to scrub generic "AI slop" from the scholar profiles.

Social Megaphone

A massive data lake containing millions of personal messages and AI chat logs used for insight extraction.

Automate the publishing pipeline for Digital Humanities insights.
Treat personal chat logs as an archaeological site for intellectual autoethnography.

The extraction from "Archive" to "Social Post" is entirely manual. The insights stay buried unless manually queried.
Signal-to-noise ratio is too high for ad-hoc searching.

The system needs persistent topic-modeling visualizations to graph intellectual focus over time, moving beyond simple keyword extraction.

Build the automated "Social Megaphone" pipeline to scan recent chats, auto-draft posts into the specific voice, and stage them in a `social_drafts` table.
Pipe the Vibe Coding Garage Discord bot directly into this database to graph community sentiment.