DEV Community

Harper Zhu profile picture

Harper Zhu

Data engineer building scalable pipelines.

Location Austin, TX Joined Joined on 
Split the Clock and Repeat the Spike

Split the Clock and Repeat the Spike

Comments
7 min read
Give Your Spike an Exit Code, Not a Retrospective

Give Your Spike an Exit Code, Not a Retrospective

Comments
6 min read
A Hunk Map for a Ninety-Minute Spike

A Hunk Map for a Ninety-Minute Spike

Comments
7 min read
Fingerprint Repeated Failures in a Ninety-Minute Spike

Fingerprint Repeated Failures in a Ninety-Minute Spike

Comments
7 min read
Ship Only a Closed Test Loop

Ship Only a Closed Test Loop

Comments
7 min read
A Write Fence for Time-Boxed Agent Spikes

A Write Fence for Time-Boxed Agent Spikes

Comments
8 min read
One Hypothesis File for a Ninety-Minute Spike

One Hypothesis File for a Ninety-Minute Spike

Comments
8 min read
Unverified Assumptions End the Spike

Unverified Assumptions End the Spike

Comments
6 min read
The Host-Receipt Gate for Agent Spikes

The Host-Receipt Gate for Agent Spikes

Comments
7 min read
Kill the Spike If the Agent Invents Config

Kill the Spike If the Agent Invents Config

Comments
6 min read
Persist AI Agent State on Free Servers with JSON Checkpoints

Persist AI Agent State on Free Servers with JSON Checkpoints

Comments
6 min read
Scoring Review Surface in a Ninety-Minute Spike

Scoring Review Surface in a Ninety-Minute Spike

Comments
6 min read
The 90-Minute Free-Server Honesty Check for AI Agents

The 90-Minute Free-Server Honesty Check for AI Agents

Comments 1
4 min read
The Token Ledger: Catching Cost Drift in AI-Generated Patches

The Token Ledger: Catching Cost Drift in AI-Generated Patches

Comments
5 min read
Pick an AI Model the Way You Pick a Dependency

Pick an AI Model the Way You Pick a Dependency

Comments
5 min read
Treat Free Tier Like a Hypothesis, Not a Feature

Treat Free Tier Like a Hypothesis, Not a Feature

Comments
4 min read
Free Models Aren't the Risk. Your Task Selection Is.

Free Models Aren't the Risk. Your Task Selection Is.

Comments
4 min read
The 90-Minute Spike That Decides Whether an AI Tool Ships

The 90-Minute Spike That Decides Whether an AI Tool Ships

Comments
4 min read
The Evidence File: A Cheap Audit Trail for AI-Generated Patches

The Evidence File: A Cheap Audit Trail for AI-Generated Patches

Comments
4 min read
Sandbox First: Why Free Servers Beat Token Grants for Agent Evaluation

Sandbox First: Why Free Servers Beat Token Grants for Agent Evaluation

1
Comments
3 min read
Green Tests Can Still Be Dead Tests: Auditing Agent-Generated Suites

Green Tests Can Still Be Dead Tests: Auditing Agent-Generated Suites

Comments
5 min read
The Runtime Is the Test Fixture: An Opinionated Harness for AI Coding Agents

The Runtime Is the Test Fixture: An Opinionated Harness for AI Coding Agents

Comments
4 min read
Free Tokens Are the Wrong Metric for AI Coding Assistants

Free Tokens Are the Wrong Metric for AI Coding Assistants

Comments
4 min read
MiniMax H3 Hype? Run a Reproducible 3-Test Coding Triage

MiniMax H3 Hype? Run a Reproducible 3-Test Coding Triage

Comments
6 min read
A No-Cost Harness for Comparing Free Coding-Agent Models and Runtimes

A No-Cost Harness for Comparing Free Coding-Agent Models and Runtimes

Comments
5 min read
Red-Teaming AI Coding Agents Without a Budget: A Boundary Test Suite on Free Models

Red-Teaming AI Coding Agents Without a Budget: A Boundary Test Suite on Free Models

Comments
6 min read
Canary Tests for AI Coding Agents: A Sandbox Harness You Can Run Yourself

Canary Tests for AI Coding Agents: A Sandbox Harness You Can Run Yourself

Comments
4 min read
A Reproducible Bug-Fix Trial for AI Coding Assistants Using Disposable Git Worktrees

A Reproducible Bug-Fix Trial for AI Coding Assistants Using Disposable Git Worktrees

1
Comments 1
5 min read
loading...