Skip to content
Directory
GitHub Star

Architecture and Systems Thinking ​

Explore ownership, authorization, reliability and evolution across modules—and whether a change improves the result users receive.

These reading groups are not a product-release checklist. Each article distinguishes design, bounded verification and actual delivery. Links marked “Chinese original” lead to articles without a published English edition.

IN THIS CATEGORY60 articles

Counts include articles in this category only; related reading is listed separately.

Tags · match all selected
01 / 05

Deep and Continuous Research

Long-running background research, staged delivery, recovery and continuity of investment theses.

  1. Agent Client Interaction Panorama: From Asking a Question to Understanding Results and Continuing Research
  2. Long-Horizon Research Panorama: Continuing Across Context, Interruptions and Time
  3. Designing Half-Hour Agent Research: From Ordinary Research to Reliable Delivery
  4. Keep Chatting: Adding Async Background Research to an Agent in a Day
  5. From Investment-Research Chat to Ongoing Research: Lessons from DeepSeek Super-agent
  6. After the Weekly Report, Does the Agent Remember Why?
  7. Why Saved HTML Can Still Fail Delivery
  8. Failover: Change the Route Without Resetting the Task
  9. A Disconnected Browser Is Not a Crashed Agent
  10. Run Termination, Stage Budgets and Safe Recovery
  11. A Research Agent Needs More Than Tools: Model, Harness, and Loop
  12. Deep Research: From Budget Upgrades to Delivery—What We Built and What the Tests Show
  13. Adaptive Research and Python
  14. Async Research: Delivering on Time After Users Leave
  15. Keeping Progressive Answers Coherent
02 / 05

Subagents and Experts

Investment perspectives, bounded peer dialogues and collaboration across expert agents.

  1. More Than One Investment Expert: How Multi-Perspective Agents Test an Investment Thesis
  2. When Two Agents Talk Directly: Bounded Bull–Bear Research
  3. Subagent Panorama: Independent Context, Specialist Work and Real Handoffs
03 / 05

Financial Research and Business Accountability

Financial data and calculation semantics, research-service boundaries and investment-business responsibilities.

  1. After General-Purpose Agent Frameworks, Why Build Financial Domain Capabilities?
  2. From a General Harness to Financial Research: A Four-Way Architecture Comparison of ETF Agent
  3. Financial Agents in Context: Systems, Lifecycles, and Accountability
  4. Trusted Handoffs Between Financial Business Agents: Why a Router Is Not Enough
  5. Why Financial Market MCP Became a Decision System
  6. Boundaries Between Research, Advice and Trading
  7. Financial Data: Time, Definitions and Reproducibility
  8. Why Python Results Can Still Be Misused in Reports
  9. Why Can a Fund Research Report Still Make Unsupported Conclusions?
04 / 05

Architecture Learning Path

An ordered getting-started series: concepts, execution, system design, hands-on research, roles, review and the six-layer architecture.

  1. Agents in 15 Minutes
  2. Prompt, Context, RAG and Memory — Where Does the Model Input Come From?
  3. Agent Intermediate: Ten Core Modules in Open ETF Agent
  4. Tools, Function Calling, MCP and Skills — How Does an Action Happen?
  5. History, Context, Memory and Checkpoints — How Does an Agent Continue?
  6. Agent Execution Identity Panorama: From a Conversation to Model Turns—What Actually Continues?
  7. Agent Deep Dive: From One Report to Durable Execution and Specialist Collaboration
  8. Bring Your Own Documents to an Agent: From the First Report to a Resumable Research Project
  9. Expert, Subagent, Review, Judge: If They All Use Another Model, Why Are They Different?
  10. After the report: source checks, specialist work, and independent review
  11. Six-Layer Request Lifecycle: From Prompt to Visible Answer
  12. Agent Context Governance Panorama · Principles: Keep Every Decision Aligned with the Goal
  13. Agent Context Governance Panorama · Practice: From the First Request to Long-Term Memory
05 / 05

System Boundaries and Evolution

Cross-module evidence, responsibility, failure attribution, reliability and long-term architecture evolution.

  1. ETF Agent vs LangGraph: Concept Mapping, Recovery, and Research Delivery
  2. From Chat to Copilot to Agent: What Actually Changes in the Architecture?
  3. Why Split an Agent into Three Processes? Independent Lifecycles Without Splitting Business Truth
  4. How Agent Development Differs from Backend Development: From One Request to a Long-Term Personal Assistant
  5. Nine Architecture Judgments Behind One Report
  6. What to Trust in Agent Engineering: From Local Facts to System Commitments
  7. Agent Fragility and Antifragility: When Safeguards Start Causing Failures
  8. The Failure Is Already Known. Why Hasn't the Call Returned? Cleanup and Result Ownership
  9. Don't Turn the Model into a State Machine
  10. Responsibilities of Model, Runtime, MCP and Skills
  11. A Trustworthy Evidence Chain for Financial Agents
  12. Why AI Agents Need Thinking, Fast and Slow
  13. Designing Agent Citations: From Host-Issued References to Revision-Safe Independent Review
  14. Code Is Cheaper; Why Are Systems Harder to Evolve?
  15. What Should Stay Stable—and What Should Change—in an Agent Architecture?
  16. Stop Blaming Every Failure on the Model
  17. Local Improvements Are Not System Progress
  18. Agent Observability Panorama: From Model Requests to Verifiable Research Facts
  19. What Does an Agent-Generated Report Actually Cost?
  20. Observing the Tool-to-Model Input Lifecycle

Modules · Engineering · Finance · Panorama · All articles

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