|
| 1 | +# Systems Development Framework |
| 2 | + |
| 3 | +Visioning Lab has developed a Systems Development framework to help organisations work through complex problems in a structured, transparent way. |
| 4 | + |
| 5 | +The approach moves from defining inputs through to action, enabling clearer thinking, better decisions and ongoing system insight. |
| 6 | + |
| 7 | +## Framework Overview |
| 8 | +Define → Structure → Insight → Action |
| 9 | + |
| 10 | +## Define |
| 11 | + |
| 12 | +Capture and clarify what matters across multiple sources and perspectives. |
| 13 | + |
| 14 | +- Documents, meetings, data, stakeholder inputs |
| 15 | +- Structured into clear points |
| 16 | +- Identification of assumptions and implications |
| 17 | + |
| 18 | +## Structure |
| 19 | + |
| 20 | +Organise relationships, options and decision logic. |
| 21 | + |
| 22 | +- Linking points into models |
| 23 | +- Defining options and trade-offs |
| 24 | +- Making reasoning explicit and comparable |
| 25 | + |
| 26 | +## Insight |
| 27 | + |
| 28 | +Identify patterns, signals and system-level implications. |
| 29 | + |
| 30 | +- Recurring issues across inputs |
| 31 | +- Hidden assumptions |
| 32 | +- Emerging risks and opportunities |
| 33 | + |
| 34 | +## Action |
| 35 | + |
| 36 | +Translate insight into interventions and next steps. |
| 37 | + |
| 38 | +- Decisions and action plans |
| 39 | +- Policy or operational changes |
| 40 | +- Feedback into future cycles |
| 41 | + |
| 42 | +## What this enables |
| 43 | +- Structured understanding of complex systems |
| 44 | +- Transparent and explainable decision-making |
| 45 | +- Comparability across cases and organisations |
| 46 | +- Continuous learning over time |
| 47 | + |
| 48 | +## Interactive Tool |
| 49 | + |
| 50 | +An interactive Next.js implementation of the framework lives in this repo. |
| 51 | +It walks the user through Define → Structure → Insight → Action, with |
| 52 | +LLM-assisted extraction at each step and editable outputs throughout. |
| 53 | + |
| 54 | +### Setup |
| 55 | + |
| 56 | +```bash |
| 57 | +npm install |
| 58 | +cp .env.example .env.local # add your ANTHROPIC_API_KEY |
| 59 | +npm run dev |
| 60 | +``` |
| 61 | + |
| 62 | +Open [http://localhost:3000](http://localhost:3000). |
| 63 | + |
| 64 | +### Stack |
| 65 | + |
| 66 | +- Next.js 14 (App Router) + TypeScript + Tailwind |
| 67 | +- Anthropic Claude (`claude-sonnet-4-6` by default) for extraction & generation |
| 68 | +- `zod` schemas for the Point / Option / Insight / Action data model |
| 69 | +- `jspdf` for the PDF export |
| 70 | +- Session state autosaved to `localStorage` (no server storage) |
| 71 | + |
| 72 | +### Project layout |
| 73 | + |
| 74 | +- [app/page.tsx](app/page.tsx) — landing |
| 75 | +- [app/tool/](app/tool/) — the four-step flow + export page |
| 76 | +- [app/api/](app/api/) — LLM endpoints (one per step) |
| 77 | +- [lib/schema.ts](lib/schema.ts) — data model |
| 78 | +- [lib/prompts.ts](lib/prompts.ts) — system prompts per step |
| 79 | +- [lib/anthropic.ts](lib/anthropic.ts) — Claude client + JSON helper |
| 80 | +- [lib/session.tsx](lib/session.tsx) — React context + localStorage autosave |
| 81 | +- [lib/pdf.ts](lib/pdf.ts) — PDF + JSON export |
| 82 | + |
| 83 | +## Applications |
| 84 | +- Infrastructure and environmental systems (e.g. water, energy) |
| 85 | +- Policy and regulatory decision-making |
| 86 | +- Research and interdisciplinary projects |
| 87 | +- Complex programme design |
| 88 | + |
| 89 | +## Relationship to Existing Work |
| 90 | + |
| 91 | +This project builds on Visioning Lab’s previous work, including: |
| 92 | + |
| 93 | +- Ontology Maker (structured terminology and relationships) |
| 94 | +- Water Decisions (decision support for infrastructure systems) |
| 95 | + |
| 96 | +It brings these together into a single, integrated framework. |
0 commit comments