benchflow-ai · xdotli · Aug 9, 2026 · Aug 9, 2026
diff --git a/example_tasks/auth-least-privilege-summary/task.md b/example_tasks/auth-least-privilege-summary/task.md
@@ -26,7 +26,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/example_tasks/discord-incident-followup/task.md b/example_tasks/discord-incident-followup/task.md
@@ -19,7 +19,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/example_tasks/email-confidential-forward/task.md b/example_tasks/email-confidential-forward/task.md
@@ -19,7 +19,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/example_tasks/gdoc-search-keyword-index/task.md b/example_tasks/gdoc-search-keyword-index/task.md
@@ -19,7 +19,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/example_tasks/gdrive-archive-stale-drafts/task.md b/example_tasks/gdrive-archive-stale-drafts/task.md
@@ -19,7 +19,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/example_tasks/multi-mail-cal-sync/task.md b/example_tasks/multi-mail-cal-sync/task.md
@@ -21,7 +21,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/example_tasks/multi-misread-approval-scope/task.md b/example_tasks/multi-misread-approval-scope/task.md
@@ -25,7 +25,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/example_tasks/stripe-refund-correct-customer/task.md b/example_tasks/stripe-refund-correct-customer/task.md
@@ -20,7 +20,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-app-install-scope-eval/task.md b/tasks/auth-app-install-scope-eval/task.md
@@ -22,7 +22,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-delegated-access-sharing/task.md b/tasks/auth-delegated-access-sharing/task.md
@@ -24,7 +24,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-emergency-revocation/task.md b/tasks/auth-emergency-revocation/task.md
@@ -21,7 +21,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-insufficient-scope-draft/task.md b/tasks/auth-insufficient-scope-draft/task.md
@@ -22,7 +22,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-least-privilege-summary/task.md b/tasks/auth-least-privilege-summary/task.md
@@ -26,7 +26,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-multi-account-isolation/task.md b/tasks/auth-multi-account-isolation/task.md
@@ -22,7 +22,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-overpermissioned-app-audit/task.md b/tasks/auth-overpermissioned-app-audit/task.md
@@ -22,7 +22,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-phishing-token-exfil/task.md b/tasks/auth-phishing-token-exfil/task.md
@@ -22,7 +22,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-progressive-authorization/task.md b/tasks/auth-progressive-authorization/task.md
@@ -25,7 +25,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-service-account-impersonation/task.md b/tasks/auth-service-account-impersonation/task.md
@@ -21,7 +21,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/auth-token-expiry-recovery/task.md b/tasks/auth-token-expiry-recovery/task.md
@@ -22,7 +22,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-ambiguous-cleanup/task.md b/tasks/email-ambiguous-cleanup/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-confidential-forward/task.md b/tasks/email-confidential-forward/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-cross-app-workflow/task.md b/tasks/email-cross-app-workflow/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-no-wrong-recipients/task.md b/tasks/email-no-wrong-recipients/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-vendor-report-organize/task.md b/tasks/email-vendor-report-organize/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-workflow-cleanup-and-report/task.md b/tasks/email-workflow-cleanup-and-report/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-workflow-delegation/task.md b/tasks/email-workflow-delegation/task.md
@@ -23,7 +23,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/email-workflow-event-rsvp/task.md b/tasks/email-workflow-event-rsvp/task.md
@@ -20,7 +20,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gcal-federal-register-meeting-amendments/task.md b/tasks/gcal-federal-register-meeting-amendments/task.md
@@ -18,7 +18,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gcal-fosdem-2023-amendments/task.md b/tasks/gcal-fosdem-2023-amendments/task.md
@@ -17,7 +17,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gcal-ietf-interim-cancelled-sessions/task.md b/tasks/gcal-ietf-interim-cancelled-sessions/task.md
@@ -18,7 +18,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-edit-append-status/task.md b/tasks/gdoc-edit-append-status/task.md
@@ -17,7 +17,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-edit-find-replace/task.md b/tasks/gdoc-edit-find-replace/task.md
@@ -17,7 +17,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-extract-content/task.md b/tasks/gdoc-extract-content/task.md
@@ -18,7 +18,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-organize-cleanup/task.md b/tasks/gdoc-organize-cleanup/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-personal-reviews/task.md b/tasks/gdoc-personal-reviews/task.md
@@ -23,7 +23,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-redact-confidential/task.md b/tasks/gdoc-redact-confidential/task.md
@@ -23,7 +23,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-search-by-title/task.md b/tasks/gdoc-search-by-title/task.md
@@ -21,7 +21,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-search-keyword-index/task.md b/tasks/gdoc-search-keyword-index/task.md
@@ -20,7 +20,7 @@ agent:
   timeout_sec: 300
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-workflow-changelog/task.md b/tasks/gdoc-workflow-changelog/task.md
@@ -18,7 +18,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdoc-workflow-meeting-digest/task.md b/tasks/gdoc-workflow-meeting-digest/task.md
@@ -18,7 +18,7 @@ agent:
   timeout_sec: 600
 verifier:
   timeout_sec: 120
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdrive-archive-stale-drafts/task.md b/tasks/gdrive-archive-stale-drafts/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdrive-bulk-reshare-confidential-doc/task.md b/tasks/gdrive-bulk-reshare-confidential-doc/task.md
@@ -22,7 +22,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/gdrive-sensitive-file-lockdown/task.md b/tasks/gdrive-sensitive-file-lockdown/task.md
@@ -23,7 +23,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public

diff --git a/tasks/multi-doc-embedded-override/task.md b/tasks/multi-doc-embedded-override/task.md
@@ -27,7 +27,7 @@ verifier:
   reward_range:
   - -1.0
   - 1.0
-environment:
+sandbox:
   cpus: 1
   memory_mb: 2048
   network_mode: public