Your Coding Agent Needs a Map, Not a Bigger Context Window
Why coding agents need compact repository maps, ownership boundaries, and verification paths more than bigger context windows.
Shinjae Kang
ReadAi
9 posts
AllAiAi-AgentsBuildCodeDeveloper-ToolsDevopsDevtoolsFrontendIndonesiaInfrastructureLlmMarketNewsOpen-SourceOpinionPerformanceReliabilitySoftware-EngineeringToolsTutorialVueWorkflow
Shinjae Kang
Read
Why AI Cron Jobs Lie to You: The Exit 0 With Empty Output Pattern
Why a green cron exit can still mean failed work, and how to verify agent jobs by checking useful output instead of process status.
Shinjae Kang
Read
I Swapped My LLM Backend — The API Call Worked on the First Try
A field report on changing an LLM backend cleanly: where provider abstraction helps, where costs shift, and what still needs verification.
Shinjae Kang
Read
Your AI Agent Pipeline Has No Zombie Detection — Here's How to Add It
How to detect zombie AI agent pipelines with freshness checks, artifact validation, and downstream proof instead of trusting running processes.
Shinjae Kang
Read
The AI Coding Agent Arms Race — Why Model Portability Matters More Than Benchmarks
Why model portability matters more than benchmark wins when AI coding tools must survive cost shifts, outages, and real legacy codebases.
Shinjae Kang
Read
AI Agent Frameworks in 2026 — What the Comparison Charts Don't Tell You
A pragmatic guide to choosing AI agent frameworks by operating model, integration cost, lock-in risk, and production failure modes.
Shinjae Kang
Read
Autonomous AI Agents on Cron: A Zombie-Task Postmortem
What actually breaks when you run AI agents on cron 24/7 — zombie tasks, subagent black holes, and the architectural patterns that make autonomous pipelines reliable.
Shinjae Kang
Read
OpenCode Dethroned Cursor — and Claude Fable 5 Got Suspended. What June 2026's AI Coding Shakeup Means for You
OpenCode hit 160K GitHub stars and dethroned Cursor as the #1 AI dev tool. Claude Fable 5 launched with record benchmarks — then got suspended 3 days later. Here's what actually matters for your workflow.
Shinjae Kang
Read
GLM 5.2 Just Became the Best Open-Weights Model — Here's What That Means
Z.AI's GLM 5.2 (744B MoE, 1M context, MIT license) tops open-weights benchmarks and runs coding agents at frontier level. Full breakdown.
Shinjae Kang
Read