Project 01 · Personal
DailyAnomaly
An autonomous website that generates and publishes a new fictional encyclopedia entry every day, with no manual step between idea and live page.
01 Context
DailyAnomaly is an autonomous generative website that publishes a new fictional encyclopedia entry every day. Each page is produced from AI generation, deterministic randomness and a semantic layout engine, so entries stay unique while remaining structurally consistent, a continuously growing archive of anomalies from a parallel knowledge system.
02 Approach
The system combines AI-driven generation with automated deployment: content is generated as structured JSON, turned into a themed static page, committed to version control, and pulled to the server, end to end, on a daily schedule.
- Daily generation: a new entry is produced and published automatically, without manual intervention.
- Dynamic identity: each page derives its layout, typography and colour from the entry's semantic metadata.
- Seed-based reproducibility: pages can be regenerated or referenced through deterministic seeds.
- Archive: past anomalies are preserved as static pages for long-term access.
- Automated pipeline: generation to live deployment runs with zero manual steps.
03 System architecture
DailyAnomaly is built as a distributed system combining AI services, automation workflows, version control and server infrastructure.
| Component | Responsibility |
|---|---|
| AI generation layer | Generates structured JSON encyclopedia entries using OpenAI models. |
| Automation engine (n8n) | Orchestrates daily content generation, formatting and publishing. |
| Semantic CSS engine | Builds page layouts and themes from metadata and anomaly level. |
| Version control (GitHub) | Stores generated pages and acts as the deployment source. |
| Deployment pipeline | Pulls updates to a VPS automatically using cron and Git. |
| Web server (Nginx) | Serves the generated pages under the DailyAnomaly domain. |