DEV Community

Finley Zhou profile picture

Finley Zhou

Backend developer who loves Rust and distributed systems.

Location Tokyo, Japan Joined Joined on 
Fixtures Are Part of the Claim: Gating Agent Patches on Test Support Diffs

Fixtures Are Part of the Claim: Gating Agent Patches on Test Support Diffs

1
Comments 1
6 min read
Route Agent Patches Before You Gate Them: Four Lanes, Fixture Provenance, and a Freeze That Needs Base-Tree Evidence

Route Agent Patches Before You Gate Them: Four Lanes, Fixture Provenance, and a Freeze That Needs Base-Tree Evidence

Comments
6 min read
Pin the Box Before You Trust the Verdict: Environment Fingerprints for Agent Patch Gates

Pin the Box Before You Trust the Verdict: Environment Fingerprints for Agent Patch Gates

Comments
6 min read
Model-Drafted Property Checks Need an Admission Test: Mutants, Fixture Hashes, and a Flaky Floor

Model-Drafted Property Checks Need an Admission Test: Mutants, Fixture Hashes, and a Flaky Floor

Comments
6 min read
When Property, Fixture, and Flake Signals Disagree: A Precedence Table for Agent Patch Gates

When Property, Fixture, and Flake Signals Disagree: A Precedence Table for Agent Patch Gates

Comments
6 min read
A Freeze Is a Ceiling on Evidence, Not a Floor: Priority Rules for Agent-Patch Test Gates

A Freeze Is a Ceiling on Evidence, Not a Floor: Priority Rules for Agent-Patch Test Gates

Comments
6 min read
Runtime Is a Budget: Ranking Property Checks by Detection per Second

Runtime Is a Budget: Ranking Property Checks by Detection per Second

Comments
7 min read
Agent Patches Can Shrink Your Test Suite Without Failing CI

Agent Patches Can Shrink Your Test Suite Without Failing CI

Comments
7 min read
Rerun Until Green Is Selection Bias: A Sequential Test for Agent-Patch Triage

Rerun Until Green Is Selection Bias: A Sequential Test for Agent-Patch Triage

Comments
6 min read
Coverage Is a Claim, Not Evidence: A Hunk-Level Ledger for Agent Patches

Coverage Is a Claim, Not Evidence: A Hunk-Level Ledger for Agent Patches

Comments
6 min read
Quarantine Is Not a Verdict: A Flake-Freeze Ledger for Agent Patch Gates

Quarantine Is Not a Verdict: A Flake-Freeze Ledger for Agent Patch Gates

Comments
5 min read
Selecting Tests for Agent Patch Queues: A Precision Budget With a Hard Floor

Selecting Tests for Agent Patch Queues: A Precision Budget With a Hard Floor

Comments
6 min read
Every Frozen Test Casts a Shadow: Auditing What Your Patch Gate Stops Seeing

Every Frozen Test Casts a Shadow: Auditing What Your Patch Gate Stops Seeing

Comments
6 min read
The Patch Edited the Fixture, Not the Code: Catching Fixture Laundering in Agent Patches

The Patch Edited the Fixture, Not the Code: Catching Fixture Laundering in Agent Patches

Comments
7 min read
Your Safeguards Can Cancel Each Other Out: A Self-Test Matrix for Agent-Patch Gates

Your Safeguards Can Cancel Each Other Out: A Self-Test Matrix for Agent-Patch Gates

Comments
5 min read
Fixture Leases: Catch Agent Patches That Pass Because Another Test Ran First

Fixture Leases: Catch Agent Patches That Pass Because Another Test Ran First

Comments
6 min read
Fixtures Rot Quietly: A Two-Corpus Audit for Agent Patches

Fixtures Rot Quietly: A Two-Corpus Audit for Agent Patches

Comments
6 min read
When Three Gates Disagree: A Precedence Router for Agent-Patch Evidence

When Three Gates Disagree: A Precedence Router for Agent-Patch Evidence

Comments
6 min read
Two Authors, One Patch: Don't Let the Agent That Wrote the Code Write the Oracle

Two Authors, One Patch: Don't Let the Agent That Wrote the Code Write the Oracle

Comments
7 min read
Your Fixtures Are the Spec: Fixture Mutation as the Missing Gate for Agent Patches

Your Fixtures Are the Spec: Fixture Mutation as the Missing Gate for Agent Patches

Comments
6 min read
Skip Is Not a Freeze: Revocable Seeded Freezes for Agent Patches

Skip Is Not a Freeze: Revocable Seeded Freezes for Agent Patches

Comments
6 min read
Budget the Gate, Not Just the Tests: Stage Caps and a Calibration Check for Agent Patches

Budget the Gate, Not Just the Tests: Stage Caps and a Calibration Check for Agent Patches

Comments
6 min read
Stop Writing Test Strategies for Agent Patches. Write a Contract Your CI Can Enforce.

Stop Writing Test Strategies for Agent Patches. Write a Contract Your CI Can Enforce.

Comments 1
7 min read
Run Less, Prove More: Diff-Scoped Test Gating for Agent Patches

Run Less, Prove More: Diff-Scoped Test Gating for Agent Patches

Comments
6 min read
A Green Run Is a Claim, Not Evidence: Re-deriving Agent Patch Test Runs

A Green Run Is a Claim, Not Evidence: Re-deriving Agent Patch Test Runs

Comments
5 min read
Property Checks Need Enemies: A Degenerate-Oracle Battery for Agent Patches

Property Checks Need Enemies: A Degenerate-Oracle Battery for Agent Patches

Comments 1
6 min read
Route Agent Patches to the Cheapest Gate That Can Catch Them

Route Agent Patches to the Cheapest Gate That Can Catch Them

Comments
7 min read
Quarantine Withdraws Coverage: Find the Only Witness Before You Freeze a Test

Quarantine Withdraws Coverage: Find the Only Witness Before You Freeze a Test

Comments
6 min read
Mutate the Fixture, Not the Code: Catching Agent Tests That Only Mirror the Implementation

Mutate the Fixture, Not the Code: Catching Agent Tests That Only Mirror the Implementation

Comments
6 min read
Three Green Reruns Prove Almost Nothing: Sizing a Flake-Detection Budget for Agent Patches

Three Green Reruns Prove Almost Nothing: Sizing a Flake-Detection Budget for Agent Patches

Comments
7 min read
Inherited Green: Freeze Debt Accounting for Agent Patches

Inherited Green: Freeze Debt Accounting for Agent Patches

Comments
7 min read
Passing Alone Is Not Passing: An Order-Sensitivity Gate for Agent Patches

Passing Alone Is Not Passing: An Order-Sensitivity Gate for Agent Patches

Comments
7 min read
Order Beats Randomness: Bisecting Test Pollution in Agent Patches

Order Beats Randomness: Bisecting Test Pollution in Agent Patches

Comments
5 min read
When a Flake Freeze Expires: Three Ledger Checks Before the Test Comes Back

When a Flake Freeze Expires: Three Ledger Checks Before the Test Comes Back

Comments
6 min read
Trusted Coverage: Stop Counting Tests You've Already Quarantined

Trusted Coverage: Stop Counting Tests You've Already Quarantined

Comments
7 min read
A Frozen Test Still Needs a Failure Path: Canaries and Exit Criteria for Quarantine Lanes

A Frozen Test Still Needs a Failure Path: Canaries and Exit Criteria for Quarantine Lanes

Comments 1
7 min read
A Freeze Needs a Witness: Seed-Matrix Triage Before You Quarantine a Flaky Test

A Freeze Needs a Witness: Seed-Matrix Triage Before You Quarantine a Flaky Test

Comments
6 min read
Did the Patch Cause the Flake, or Find It? A Base/Head Attribution Gate for Agent Patches

Did the Patch Cause the Flake, or Find It? A Base/Head Attribution Gate for Agent Patches

Comments
6 min read
Quarantine Is a Loan, Not a Fix: Attributing Agent-Patch Flakes Before You Freeze Them

Quarantine Is a Loan, Not a Fix: Attributing Agent-Patch Flakes Before You Freeze Them

Comments
7 min read
One Manifest Diff Catches Agents Weakening Tests, Fixtures, and Flake Freezes

One Manifest Diff Catches Agents Weakening Tests, Fixtures, and Flake Freezes

Comments
6 min read
Patches Are Cheap Now. Verification Isn't: A Tiered Evidence Budget for Agent Patches

Patches Are Cheap Now. Verification Isn't: A Tiered Evidence Budget for Agent Patches

Comments
6 min read
Order Is a Hidden Input: A Permutation Verdict for Agent Patches

Order Is a Hidden Input: A Permutation Verdict for Agent Patches

Comments
7 min read
Fewer Tests Ran, All Tests Passed: A Collection Census for Agent Patches

Fewer Tests Ran, All Tests Passed: A Collection Census for Agent Patches

Comments
6 min read
Agent Patches Don't Need the Whole Suite: A Skip-Witness Gate With Patch-Keyed Shadow Sampling

Agent Patches Don't Need the Whole Suite: A Skip-Witness Gate With Patch-Keyed Shadow Sampling

Comments
7 min read
The Agent Patch Loosened the Assertion and the Suite Stayed Green: An Oracle-Latitude Gate

The Agent Patch Loosened the Assertion and the Suite Stayed Green: An Oracle-Latitude Gate

Comments
6 min read
Which Gate Should Run First? Ordering Agent-Patch Checks by Yield per Second

Which Gate Should Run First? Ordering Agent-Patch Checks by Yield per Second

Comments
6 min read
The Assertion Got Weaker and the Build Went Green: Diffing Oracle Strength in Agent Patches

The Assertion Got Weaker and the Build Went Green: Diffing Oracle Strength in Agent Patches

Comments
7 min read
Retries Are Not Evidence: A Sample-Size Gate for Agent Patches

Retries Are Not Evidence: A Sample-Size Gate for Agent Patches

Comments
7 min read
Your Agent-Patch Gates Overlap. Measure It Before You Add a Fourth.

Your Agent-Patch Gates Overlap. Measure It Before You Add a Fourth.

Comments 1
7 min read
Canary Mutations for Fixtures: A Sensitivity Gate for Agent-Written Patches

Canary Mutations for Fixtures: A Sensitivity Gate for Agent-Written Patches

Comments
7 min read
Your Agent Patch Edited conftest.py: A Fixture-Contract Gate With a Blast-Radius Map

Your Agent Patch Edited conftest.py: A Fixture-Contract Gate With a Blast-Radius Map

Comments
5 min read
The Agent Patch Edited the Fixture, Not the Bug: A Provenance Gate for Test Data

The Agent Patch Edited the Fixture, Not the Bug: A Provenance Gate for Test Data

Comments
5 min read
The Fixture Moved With the Code: Pinning Test Inputs Before an Agent Patch Rewrites Its Own Oracle

The Fixture Moved With the Code: Pinning Test Inputs Before an Agent Patch Rewrites Its Own Oracle

Comments
6 min read
A Frozen Test Can't Justify a Patch: Lease-Bound Freezes and Witnessed Properties for Agent Patches

A Frozen Test Can't Justify a Patch: Lease-Bound Freezes and Witnessed Properties for Agent Patches

Comments
6 min read
The Patch Didn't Break the Test. It Added a Clock: An Ambient-Input Ledger for Agent Patches

The Patch Didn't Break the Test. It Added a Clock: An Ambient-Input Ledger for Agent Patches

Comments
6 min read
Nothing Failed. The Suite Just Got Slower: A Wall-Clock Budget Gate for Agent Patches

Nothing Failed. The Suite Just Got Slower: A Wall-Clock Budget Gate for Agent Patches

Comments 1
6 min read
The Patch Added a Sleep: Reviewing Timing and Retry Hunks Before Merge

The Patch Added a Sleep: Reviewing Timing and Retry Hunks Before Merge

Comments
6 min read
Green by Forgiveness: Three Budgets That Catch Agent Patches Passing for the Wrong Reason

Green by Forgiveness: Three Budgets That Catch Agent Patches Passing for the Wrong Reason

Comments
6 min read
A Freeze With an Expiry: Rate-Based Triage for Flaky Tests After an Agent Patch

A Freeze With an Expiry: Rate-Based Triage for Flaky Tests After an Agent Patch

Comments
6 min read
Cheapest First: Sequencing Agent-Patch Checks by Cost, and a Freeze That Re-Arms Itself

Cheapest First: Sequencing Agent-Patch Checks by Cost, and a Freeze That Re-Arms Itself

Comments
5 min read
loading...