Skip to content

ETF Agent · NotesUnderstanding agents through real engineering.

How complex agents run, why their architecture evolves, and how AI-assisted engineering delivers working systems.

ESSENTIAL READING · START HERE

01 / THE SYSTEM MAPSix-layer request lifecycle V3Follow a question through admission, context, Main and subagents, optional tools, Publication and the user interface.Read the complete architecture →02 / THE DECISION MAPNine architecture principlesUnderstand who owns semantics, permission, effects, recovery and the complexity of long-lived systems.Explore the core principles →

SELECTED ARTICLES

Fresh perspectives

Browse all articles

EXPLORE MORE

Go a little deeper

Open any section when you need it.

Choose a path by interestFrontend, Agent apps, architecture, and AI-assisted engineering
Development updates and Owner statusEvolution, evidence scope, and future work

The architecture gallery maintains the detailed Owner inventory and evidence scope. Planned writing lives in the roadmap instead of a daily changelog on this homepage.

About this site and the authorScope, contact, and collaboration

This site publishes engineering articles, architecture diagrams, and its own source code. The full Agent application, production credentials, and production metrics are not public. A local validation result is not a production guarantee.

Engineering discussion, not investment advice. Full reading paths and historical editions remain available.

Engineering notes, evidence and boundaries. · Contact: hpoenixf@foxmail.com · WeChat: hpoenixf0510