Silin-Chen's picture
CrossSWE-Bench: current 171-instance release
0e239c6
Raw History Blame Contribute Delete
2.4 kB
{
"schema_version": 1,
"instance_id": "cloudnative-docker-cli-6580",
"kind": "command_contract",
"network": "none",
"repeat_count": 1,
"required_regression_worlds": [
"all_buggy",
"cli_only",
"moby_only",
"all_fixed"
],
"test_files": [],
"specifications": [],
"commands": [
{
"repository": "cli",
"container_path": "/testbed/docker/cli",
"setup_command": null,
"environment": {},
"fail_to_pass": "cd /testbed/docker/cli && cp vendor.mod go.mod && cp vendor.sum go.sum && printf '\\nreplace github.com/moby/moby/client => /testbed/moby/moby/client\\n' >> go.mod && go test -mod=mod -json ./cli/command/image -run '^TestCross90LoadBehavior$' -count=1",
"pass_to_pass": "cd /testbed/docker/cli && cp vendor.mod go.mod && cp vendor.sum go.sum && printf '\\nreplace github.com/moby/moby/client => /testbed/moby/moby/client\\n' >> go.mod && go test -mod=mod -json ./cli/command/image -run '^TestCross90LoadRegression$' -count=1"
},
{
"repository": "moby",
"container_path": "/testbed/moby/moby",
"setup_command": null,
"environment": {},
"fail_to_pass": "cd /testbed/moby/moby/client && go test -json . -run '^TestCross90LoadContract$' -count=1",
"pass_to_pass": "cd /testbed/moby/moby/client && go test -json . -run '^TestCross90LoadRegression$' -count=1"
}
],
"execution_policy": {
"test_patches": "Apply in listed order after candidate patches, in an isolated evaluator workspace.",
"commands": "Complete source setup in setup_order, then evaluator setup in evaluation_setup_order, ready checks, pass_to_pass and fail_to_pass commands; teardown in reverse evaluator setup order.",
"world_placeholder": "A world labels the prepared source state for both regression and behavior commands: all_buggy for base, all_fixed for gold, candidate for submitted patches.",
"success": "All required behavioral and regression tests pass; infrastructure failures are not behavioral outcomes.",
"isolation": "The agent receives base sources and statement only; evaluation and reference materials remain outside its workspace."
},
"availability": {
"test_materials": "included",
"portable_environment": "pending",
"dependencies": "Some commands require external toolchains or dependency assets; complete portable recipes are a follow-up release."
}
}