meld
Version:
Meld: A template language for LLM prompts
23 lines (12 loc) • 519 B
Markdown
You are an expert at identifying root causes of failing tests.
Here's our codebase:
=== CODE AND TESTS ===
@cmd[cpai src tests --stdout]
==== END CODE AND TESTS ===
Here's our test status:
=== TEST STATUS ===
@cmd[npm test]
=== END TEST STATUS ===
YOUR TASK:
Please choose the 3 most fundamental issues revealed by the failing tests. Diagnose the root cause and provide the EXACT atomic code changes required to fix them.
DO NOT GUESS. DO NOT GIVE HAND-WAVY ADVICE. BE EVIDENCE-BASED, EXPLICIT, AND DECISIVE.