Context engineering is not a new activity. It is a discipline we have been practising for decades under different names.
Context Engineering: Designing AI That Understands, Remembers, Acts, and Learns is out now on Kindle.
Most AI systems do not fail because the model is wrong. They fail because the context is.
Configuration management gave us baselines and change control. Systems engineering gave us decision gates and requirements traceability. Domain-driven design gave us the ubiquitous language and bounded contexts. Then LLMs arrived, and much of that discipline was set aside: context became "the prompt", something to tweak rather than something to version, validate and govern.
That is the argument of the book. Context engineering is an old discipline applied to a new kind of system, one where the software is partly prompts, partly knowledge, partly retrieval, partly policy, and partly the model's emergent behaviour. Terraphim exists because of this argument: role-aware search, knowledge graphs and deterministic retrieval are context infrastructure, not features.
Across 25 chapters, the book covers:
- why AI software fails differently: context drift, not crashes
- knowledge graphs as context infrastructure, and when deterministic search beats vector embeddings
- token budgets, compression and skill architecture
- judge harnesses, typed decision models and evidence lineage: how to make an AI's decisions inspectable
- validation gates, multi-agent context and enterprise governance
- building a context engineering practice in your team
It is written for senior developers, engineering managers and architects who need AI to be dependable in production, not just impressive in a demo.
Get the book: Context Engineering on Kindle (also on Amazon UK).
