Evaluation Audit: Terminal Bench 2.1 findings

[agentecobuilder]

We are collecting public feedback on scoring errors, ambiguous items, contamination, or unfair comparisons in Terminal Bench 2.1.

Benchmark: https://stratix.layerlens.ai/benchmarks/6a29ec63e014f99a0f32ed31

Please open a new issue (or comment here) with:

• Link or ID of the specific item
• What’s wrong
• Why it matters
• Suggested fix (optional)

All real findings will be credited publicly!

Notion Info Page