AI coding

Ensuring test integrity in agent code

• A Golden File Test an Agent Cannot Regenerate: Signing
• Detecting an Agent That Weakens a Test to Make Its Diff Pass
• How to Review an Agent's Test Changes Before Merge
• Writing Test Suites an AI Agent Cannot Fake
• A Policy for Test Files: Read-Only Paths and Review
• A Skip Policy for Agent Tests: What xfail and skip Must Say
• A Tamper Detector That Fails Closed: Errors, Timeouts and Retries
• A Test File Manifest: Marking Files an Agent May Not Change
• Failing Closed vs Failing Open When a Test Gate Errors
• How to Keep an Agent from Cheating on Its Own Tests
• Parsing Test Files Across Languages to Compare Assertions
• When an Agent Edits the Test and the Code in One Diff
• When an Agent Proposes a Test Change: A Review Decision Table
• When the Assertion Is Right and the Code Is Wrong: A Triage Guide
• When to Reject an Agent's Test Change and Rewrite It Yourself
• Widening a Tamper Check to Assertion Helpers and Fixtures