Qingfeng

Building in public with local AI agents. I hand my workday to an agent that runs on my own machine - screen, shell, browser, phone over ADB. https://agent-home.ai

Joined on Oct 20, 2026

Qingfeng

Oct 21

The Two Lists I Keep for an Agent That Runs Unattended

#discuss #productivity #programming #ai

2 min read

Want to connect with Qingfeng?

Create an account to connect with Qingfeng. You can also sign in below to proceed if you already have an account.

Create Account

Already have an account? Sign in

Qingfeng

Oct 21

The Hardest Part of a Local Agent Is Deciding When to Stop

#showdev #programming #productivity #ai

2 min read

Qingfeng

Oct 21

A Script and a Task Are Not the Same Thing

#discuss #productivity #programming #ai

2 min read

Qingfeng

Oct 21

I Read My Agent's Log Backwards, and the Bug Was in Line Two

#productivity #testing #programming #ai

2 min read

Qingfeng

Oct 21

What I Automate First in a New Project, and Why It Is Never the Important Part

#devops #productivity #programming #ai

2 min read

Qingfeng

Oct 21

Letting the Agent Touch the Browser Changed My Task List

#showdev #programming #automation #ai

2 min read

Qingfeng

Oct 21

The Twenty Minute Review That Keeps My Agent Honest

#discuss #showdev #productivity #ai

2 min read

Qingfeng

Oct 21

What I Got Wrong About "Just Give It a Better Model"

#discuss #programming #productivity #ai

2 min read

Qingfeng

Oct 21

The Second Run Is Where the Lies Show Up

#discuss #testing #programming #ai

2 min read

Qingfeng

Oct 21

The 20% of Tasks I Deliberately Do Not Give the Agent

#discuss #showdev #productivity #ai

2 min read

Qingfeng

Oct 21

I Sorted 400 Agent Failures Into Five Buckets. Almost None Were the Model's Fault.

#discuss #productivity #programming #ai

2 min read

Qingfeng

Oct 21

Two Plain Text Files Replaced My Whole Agent Framework

#showdev #productivity #programming #ai

2 min read

Qingfeng

Oct 21

The Agent Doesn't Need More Context. It Needs the Right Five Lines.

#discuss #productivity #programming #ai

2 min read

Qingfeng

Oct 21

If I Started Over, I'd Write the Logs First

#showdev #devops #programming #ai

2 min read

Qingfeng

Oct 21

My Agent Was Never Slow. My Tasks Were Serial.

#showdev #productivity #programming #ai

2 min read

Qingfeng

Oct 21

What I Do When the Agent Says It's Done

#showdev #productivity #programming #ai

2 min read

Qingfeng

Oct 21

The Pre-Flight Check That Killed Most of My Failures

#showdev #devops #programming #ai

2 min read

Qingfeng

Oct 21

Five Things I Check Before Believing an Agent Demo

#showdev #productivity #discuss #ai

2 min read

Qingfeng

Oct 21

The Ten Unglamorous Tasks My Agent Actually Handles

#discuss #productivity #showdev #ai

2 min read

Qingfeng

Oct 20

What I Let the Agent Do Without Asking, and What I Never Will

#showdev #productivity #security #ai

2 min read

Qingfeng

Oct 20

I Stopped Writing Prompts and Started Writing Fixtures

#showdev #testing #productivity #ai

2 min read

Qingfeng

Oct 20

Cleaning Up After a Failed Run Is the Part Nobody Writes About

#showdev #devops #programming #ai

2 min read

Qingfeng

Oct 20

How I Measure Whether My Agent Is Actually Saving Me Time

#discuss #showdev #productivity #ai

2 min read

Qingfeng

Oct 20

The Log Format That Made My Agent Debuggable

#showdev #productivity #programming #ai

2 min read

Qingfeng

Oct 20

Coordinate Clicks Are a Trap: Driving a GUI Agent Without Pixel Guessing

#showdev #automation #testing #ai

2 min read

Qingfeng

Oct 20

The Approval Gate: Designing an Agent That Can't Burn Your House Down

#showdev #ai #security #agents

2 min read

Qingfeng

Oct 20

Giving an AI Agent a Real Phone: What ADB Automation Actually Looks Like

#showdev #android #adb #automation

2 min read

Qingfeng

Oct 20

30 Days of a Local AI Agent on My Own Machine: What Actually Broke

#showdev #ai #llm #agents

2 min read

DEV Community

Qingfeng

Badges

Writing Debut

The Two Lists I Keep for an Agent That Runs Unattended

Want to connect with Qingfeng?

The Hardest Part of a Local Agent Is Deciding When to Stop

A Script and a Task Are Not the Same Thing

I Read My Agent's Log Backwards, and the Bug Was in Line Two

What I Automate First in a New Project, and Why It Is Never the Important Part

Letting the Agent Touch the Browser Changed My Task List

The Twenty Minute Review That Keeps My Agent Honest

What I Got Wrong About "Just Give It a Better Model"

The Second Run Is Where the Lies Show Up

The 20% of Tasks I Deliberately Do Not Give the Agent

I Sorted 400 Agent Failures Into Five Buckets. Almost None Were the Model's Fault.

Two Plain Text Files Replaced My Whole Agent Framework

The Agent Doesn't Need More Context. It Needs the Right Five Lines.

If I Started Over, I'd Write the Logs First

My Agent Was Never Slow. My Tasks Were Serial.

What I Do When the Agent Says It's Done

The Pre-Flight Check That Killed Most of My Failures

Five Things I Check Before Believing an Agent Demo

The Ten Unglamorous Tasks My Agent Actually Handles

What I Let the Agent Do Without Asking, and What I Never Will

I Stopped Writing Prompts and Started Writing Fixtures

Cleaning Up After a Failed Run Is the Part Nobody Writes About

How I Measure Whether My Agent Is Actually Saving Me Time

The Log Format That Made My Agent Debuggable

Coordinate Clicks Are a Trap: Driving a GUI Agent Without Pixel Guessing

The Approval Gate: Designing an Agent That Can't Burn Your House Down

Giving an AI Agent a Real Phone: What ADB Automation Actually Looks Like

30 Days of a Local AI Agent on My Own Machine: What Actually Broke