Topic Hub
AI-Assisted Software Engineering
A practical map of the context, repository structure, review habits, agent boundaries, and human judgment that make AI-assisted delivery more reliable.
The visible change is faster code generation. The deeper change is that software teams must make intent, constraints, architecture, and evidence easier for both people and AI to use. This hub connects the mental models, working practices, and systems behind that shift.
Topic orientation
The real bottleneck is the engineering system around the model.
AI assistants can produce code quickly, but they do not carry a team's durable understanding by default. Useful outcomes depend on the context they receive, the boundaries they can see, and the review system that evaluates what they produce.
This body of work asks four practical questions: What context does the task need? How should a repository expose it? Which work belongs to AI? How do humans review and own the result?
Where to start
Build the mental model before adding more tools.
This short path moves from expectations, to context, to the repository as an operational knowledge system.
- Mental model
Coding Assistants Are Not Junior Developers
Start by separating a context-driven tool from the way a human teammate learns and remembers.
- Context
Stop Prompting Harder. Start Giving Better Context.
See why requirements, constraints, decisions, and examples matter more than clever prompt wording.
- Repository
How I Structure Repositories So AI Can Actually Understand Them
Turn context into durable repository structure that humans and assistants can navigate.
Explore by subtopic
Follow the question you are trying to solve.
Context beyond the prompt
Connect code context to work intent, business rules, and organizational knowledge.
Review and task allocation
Decide what AI should do and how to evaluate plausible generated work.
Agents and orchestration
Move from one assistant to bounded agents, separated roles, and explicit coordination.
Systems and evidence
See the ideas tested in a working system.
Writing explains the choices. Systems shows what those choices look like under implementation and review constraints.
Topic hub and series
Use the hub to explore. Use the series to progress in order.
This page groups ideas by question and includes independent writing and systems. The AI-Assisted Software Engineering series is a deliberate nine-part sequence for readers who want a structured path from mindset to orchestration.