Terminal-Bench 2.0 removes OpenBlocks after cheating verification

• Terminal-Bench maintainers said they "independently verified" cheating claims against OpenBlocks and removed it from the 2.0 leaderboard, reversing the earlier claim that OB-1 was "#1 on Terminal Bench" removal notice ranking claim.
• The practical takeaway for engineers is that leaderboard position alone was not durable here: the maintainers pointed to public submission artifacts as part of a community-auditable process submission note.
• Before the removal, OpenBlocks had been promoted as usable with a local model subscription, with the linked repo describing a local proxy for routing requests to OpenAI-compatible backends instead of the default cloud path local model post OB1 repo.
• A supporting implementation detail from the same thread was OB-1's "microcompact" behavior, described as compressing context during a session to preserve longer working context microcompact note.

Terminal-Bench 2.0 maintainers said they "independently verified these claims" and removed OpenBlocks from the leaderboard. The same post says recent submissions live in a public dataset, making it easier for the community to inspect entries and "detect cheating" removal notice leaderboard dataset.

That matters because OpenBlocks had just been advertised as "#1 on Terminal Bench" in the launch thread for OB-1 ranking claim. For engineering teams comparing coding agents, this is a reminder that public rankings are only as useful as the submission rules, harness transparency, and the maintainers' willingness to revoke results after audit.

Before the removal, the OpenBlocks thread paired the ranking claim with product details: OB-1 could be used with a "local model subscription" and was said to be tested with GPT-5.4 and Opus 4.6 local model post. The linked repository describes a local proxy setup that can rewrite OB-1 requests to a local API server or other OpenAI-compatible backend, letting users bring their own keys or route through infrastructure they control OB1 repo.

A follow-up post also highlighted a "microcompact" feature that reportedly compacts context during the session so the model keeps a "longer refreshed context" microcompact note. That may still be a useful implementation idea, but the leaderboard removal means engineers should separate those product mechanics from the now-invalidated benchmark standing.