fabro/evals/swe-bench/workflow.fabro
Bryan Helmkamp e246b7f4bb Add SWE-bench evaluation harness
Python scripts for running SWE-bench Lite evals against Fabro agent
in Daytona sandboxes: instance orchestration, Dockerfile generation,
and result evaluation via the official swebench harness.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-16 08:37:24 -04:00

12 lines
428 B
Text

digraph SWEBench {
rankdir=LR
start [shape=Mdiamond]
exit [shape=Msquare]
setup [label="Setup", shape=parallelogram, script="$setup_script"]
solve [label="Solve", prompt="Fix this GitHub issue in the repository. Make the minimal code change needed."]
extract_patch [label="Extract Patch", shape=parallelogram, script="git diff"]
start -> setup -> solve -> extract_patch -> exit
}