TermGrade: 1,004 open-source RL environments for terminal agents boost Terminal-Bench by +3.1
DavidBennett__ · x · 2026-10-09
Weyaxi's team released TermGrade: fully open-source RL environments for terminal agents.
- 1,004 executable environments, each verified by execution
- 36k trajectories (failures included) behind grades across 6 models
- Difficulty measured per model actually being trained
- Training gemma-4-31B-it on tasks it solved half the time yielded +3.1 on Terminal-Bench 2.1
- Full build/grading story and the RL recipe are published
The authors frame RL environments as an emerging serious data discipline for agentic RL.
More from coding & agent
- Runway MCP + Claude Opus 5.5 Demo Wows Users: "I'd Pay $120 for This" — tlakomy · 2026-10-09
- Memory Journal: A Portable .txt Format That Transfers AI Context and Skills — NoobSolid26 · 2026-10-09
- Bluerails Launches Hotel-Booking MCP Server Covering 280K Properties in DACH — bookwithbluerails · 2026-10-09
- Voyager launches as 'Codex for creative work', driving Blender, Resolve and After Effects — rohanpaul_ai · 2026-10-09
- PartyKit founder shuts down free hosted platform 2.5 years after Cloudflare acquisition — threepointone · 2026-10-09
- world-intel-mcp: open-source MCP server with 132 tools for real-time global intelligence — tom_doerr · 2026-10-09