Memdoor reconstructs a multi-agent coding workflow
A Memdoor reconstruction of the GVS5H method assigns programming tasks to five fresh-context Qwen3.8-27B roles plus a repair worker, with agents communicating through files and shell scripts deciding test success. Across five passes on the first ten LiveCodeBench Hard problems, the workflow averaged 9.8 out of 10, versus 9.6 for the authors' harness and 8.0 for single-call Fable 5 results. The report says all 50 runs, grader outputs, costs and serving hosts were published.
The reconstruction provides detailed logs for a multi-agent programming workflow that reported a 9.8/10 average on 50 LiveCodeBench Hard attempts, offering a concrete reference for agents that coordinate through disk files rather than shared chat context.
A multi-agent harness for programming problems, written as a Memdoor workflow. Five roles (planner, two ideators, solver, manager) and a repair worker each start from a fresh context and talk only through files on disk. A shell script that runs the samples decides success, not a model. It is the method of GVS5H (Lee, 2026: five Qwen3.8-27B models match Fable 5 on LiveCodeBench Hard).
Rebuilt with the authors' exact prompts on Qwen3.8-27B, their loop scored a mean of 9.8 of 10 over five passes on the first ten LiveCodeBench Hard problems, against 9.6 for their own harness and 8.0 for Fable 5 alone (below, every run published).
The coder wrote these files itself from a description of the method (2026-10-25). Nobody hand-wrote them.
Five passes on the authors' ten (2026-10-31)
GVS5H's own harness, rebuilt as a Memdoor workflow on the same model: Qwen3.8-27B, a 128K reply cap, the authors' prompts and loop. The first ten problems of their LiveCodeBench Hard list, five passes, every program graded on every public and hidden test with their corrected grader.
#Memdoor#GVS5H#Qwen3.8-27B#Fable 5#LiveCodeBench Hard#OpenRouter