PASS or FAIL?

Background

The Critic audits an answer for two kinds of problems: GAPs (sub-questions with no useful data the answer glosses over) and UNSUPPORTED claims (statements not backed by any finding). The verdict rule:

FAIL if any gap or unsupported claim was found, otherwise PASS.

Problem statement

Implement critic_verdict(gaps, unsupported) returning the verdict string.

Input

• gaps — list of gap descriptions (possibly empty).
• unsupported — list of unsupported-claim descriptions (possibly empty).

Output

Returns "FAIL" if either list is non-empty, else "PASS".

Examples

Example 1 — clean

Input:  gaps = [], unsupported = []
Output: "PASS"

Example 2 — an unsupported claim

Input:  gaps = [], unsupported = ["adopted in StarCoder and CodeLlama"]
Output: "FAIL"

Constraints

• Any non-empty list → "FAIL".
• Both empty → "PASS".

Notes

• The verdict is the machine-readable signal an orchestrator branches on — diagnosis the next lesson will finally act on.