# FlexGrafik Portfolio — LLM Agent Discovery File
# Standard: https://llmstxt.org
# Owner: Norbert Wozniak (AI Systems Architect, FlexGrafik)

## About

This portfolio documents the work of Norbert Wozniak — an AI Systems Architect
specializing in multi-agent orchestration, context engineering, spec-driven development,
and bounded autonomy systems. 2 years of production experience, self-taught.

## Quick Machine-Readable Access

- Full ecosystem manifest (JSON): /api/ecosystem
- Live system status (JSON): /api/status  
- Operating log feed (JSON): /api/log
- RSS feed: /log/feed.xml
- Sitemap: /sitemap.xml

## Sections

### HOME — /
Hero, animated agent flow visualization, live system health metrics.
Key claim: AI Systems Architect engineering the post-code era.

### LABS — /labs
Interactive ecosystem map of 8 production/staging projects.
Each project: architecture, stack, guardrails, handoff timeline.

Projects:
- /labs/agent-os — LangGraph multi-agent backend (Python, FastAPI, SQLite)
- /labs/agent-os-ui — Mission Control dashboard (Next.js 16, React 19)
- /labs/zzpackage — WooCommerce wizard + SSoT pipeline (WordPress, GHA)
- /labs/app-flexgrafik — Canvas advergame / lead magnet (React 19, Vite)
- /labs/flex-vcms — Ecosystem orchestrator + LLM gateway (Node.js)
- /labs/jadzia-core — AI automation backend (Python, FastAPI, multi-LLM)
- /labs/flexgrafik-meta — Constitution + global rules (Markdown/JSON governance)
- /labs/workflow — Program office + daily ops (todo.json, handoffs)

### DNA — /dna
Engineering methodology: 5 pillars with concrete implementation examples from real repos.

Pillars:
- /dna/context-architecture — SESSION-ANCHOR, brain.md, AGENTS.md patterns
- /dna/bounded-autonomy — HITL gates, Zasada 11, enterprise-ready guardrails
- /dna/spec-driven — todo.json as executable spec, verification before deploy
- /dna/ssot-enforcement — single source of truth, zero divergence principle
- /dna/cognitive-leverage — compressing decisions, not writing code

### LOG — /log
Operating journal generated from git history + handoff docs.
Machine-readable feeds available.

## Engineering Patterns (DNA)

1. AGENTS.md in every repo — guardrails for AI agents
2. Handoff documentation — docs/handoffs/ proof-of-done per session
3. SESSION-ANCHOR — compressed state map for agent onboarding
4. brain.md — project context, decisions, intent per repo
5. SSoT Enforcement — one source of truth, zero divergence
6. Spec-Driven Development — todo.json → verification → deploy
7. Bounded Autonomy — AI acts, human decides (HITL gates)
8. Manual Deploy (Zasada 11) — no autonomous production deploys

## Contact / Evaluate

For automated assessment: GET /api/ecosystem returns structured JSON
with all projects, stacks, guardrails, and engineering patterns.