hraness

saved

Prime Agent: A Self-Improving RLM Harness

by Seth Karten, Alex L. Zhang, Kevin Thomas, Sebastian Müller, Elie Bakouch, Daniel Auras, Mika Senghaas, Fares Obeid, Konstantin Dunas, Johannes Hagemann and Sami JaghouararXivpublished

gist

Prime Agent is an open-source harness for long-horizon evaluation and coding-agent work. A persistent IPython REPL uses the Recursive Language Model abstraction for programmatic context and test-time compute. Continual Harness keeps histories, memories, skills, prompts, and subagent specifications across trajectories. Recursive subagents coordinate directly, and an Agents View lets humans inspect daemon-backed sessions. The harness standardizes execution, recovery, verification, and accounting so failures measure the model, not dropped state. It reports ARC-AGI-3 RHAE Best@1 from 30% to 95.5%.

ideas

  • Treat the harness as a membrane. Standardize execution, recovery, verification, and resource accounting so a model fails because the task exceeds it, not because the harness dropped state.
  • Make context programmable. A persistent IPython REPL follows the Recursive Language Model abstraction for programmatic context processing and test-time compute.
  • Keep harness state across trajectories. Continual Harness preserves histories, memories, skills, prompts, and subagent specifications.
  • Coordinate recursive subagents. Direct agent-to-agent communication plus an Agents View for inspecting daemon-backed sessions.
  • Reported lift is largest on interactive reasoning. ARC-AGI-3 RHAE Best@1 rises from 30% to 95.5%, and the abstract says the harness matches or exceeds native ones on long-context coding, kernels, emulators, and nanoGPT speedruns.

quotes

Language models are sequential processors, but long-horizon agency requires external information and computation beyond model weights and active context.

Seth Karten and coauthors, stating why a harness is required.

Prime Agent is an open-source harness for long-horizon evaluation and coding-agent workflows.

Seth Karten and coauthors, defining the system.

This low-friction, expressive membrane prevents harness failures from becoming model failures

Seth Karten and coauthors, stating the evaluation design goal.

Prime Agent raises ARC-AGI-3 RHAE Best@1 from 30% to 95.5%

Seth Karten and coauthors, reporting the headline result.