Skip to content
  1. Home
  2. Benchmarks
  3. Integration Bench
  4. Gemini 3.8 Flash

Gemini 3.8 Flash

Integration capability of coding agents on Gemini 3.8 Flash.

Rank #7 of 17 Pass 1 59.68 ±1.76 Resolved 60% of tasks Harness OpenCode Effort high Total cost not reported Mean time 14 min Mean tool calls 111 Graded 50/50
64
polling
64% of tasks touching this surface resolved
62
writeback
62% of tasks touching this surface resolved
27
webhooks
27% of tasks touching this surface resolved

Published attempts

50 of 50 scored tasks released

One official run per task. Click a row for the full trajectory: every tool call, vendor request, graded check and the diff.

TaskTicketScoreChecksTool callsTimeCostOutcome
task-0001Sync StaffLine candidates, jobs, applications, and notes into our canonical store9229/30149 (17 failed)12 minresolved
task-0002Legal-hold export: TalentForge custodians and everything on file against them10064/64125 (22 failed)16 minresolved
task-0003Build the weekly Placemint contact sweep0341/34379 (15 failed)10 minpartial
task-0004Merge StaffLine application stage with Placemint placement stage048/64146 (17 failed)14 minmandatory fail
task-0005Keep TalentForge candidates fresh: webhooks + reconciliation polling020/23162 (19 failed)20 minmandatory fail
task-0006The case-note mirror is short every morning100159/15993 (14 failed)18 minresolved
task-0007Harden the Vettly sync connector for production scheduling10039/39113 (19 failed)14 minresolved
task-0008Build a StaffLine full-sync + writeback connector10033/33119 (15 failed)11 minresolved
task-0009Migrate off the legacy Bullpen app-token integration10040/40117 (11 failed)10 minresolved
task-0010Build the GlobalHire audit-period extract Compliance keeps asking us for10081/81111 (11 failed)10 minresolved
task-0011TalentForge connector: consume webhooks while pushing corrections upstream046/52132 (16 failed)16 minmandatory fail
task-0012Build the full TalentLoop connector: polling, webhooks, and writeback0102/121154 (13 failed)17 minmandatory fail
task-0013The warehouse and RecruitOS have been out of step for months10096/96103 (9 failed)11 minresolved
task-0014HireWire: clear the stale-screening correction backlog, exactly once10016/16105 (14 failed)11 minresolved
task-0015Bulk-import a migration batch into StaffLine022/26108 (11 failed)14 minmandatory fail
task-0016HireWire connector: push stage-change events + keep an incremental poll fresh9245/48140 (12 failed)25 minresolved
task-0017Move the Brightmoor candidate mirror onto TalentForge's event subscription044/156118 (14 failed)14 minpartial
task-0018Rebuild the nightly requisition rollup on the RecruitOS API100142/14277 (12 failed)10 minresolved
task-0019Rebuild the Vettly change feed for the vaultd cutover100184/184109 (13 failed)22 minresolved
task-0020Sync GlobalHire candidates into our canonical store (offset polling)10034/34112 (18 failed)12 minresolved
task-0021Sync GlobalHire candidates, placements, and agencies into our canonical store10022/2292 (12 failed)11 minresolved
task-0022Harden the GlobalHire backfill for a production tenant026/27164 (18 failed)52 minmandatory fail
task-0023TalentLoop connector: keep a candidate/application mirror deletion-accurate018/26113 (14 failed)16 minmandatory fail
task-0024TalentLoop connector: unified freshness across webhook-covered and poll-only entities027/40125 (19 failed)16 minmandatory fail
task-0025Reset the CrewCall roster watermark after the tenant rebuild0500/503105 (11 failed)10 minpartial
task-0026Push the Q1 close back into Placemint0145/26831 (1 failed)1 minpartial
task-0027Reliability: Interviewly connector loses and regresses updates10024/2497 (15 failed)17 minresolved
task-0028Harden the CrewCall snapshot connector for production10044/4485 (18 failed)10 minresolved
task-0029Get the Riverside signup file onto the CrewCall roster100102/10299 (12 failed)9 minresolved
task-0030Census the desk board against the Placemint placement book100325/32597 (12 failed)11 minresolved
task-0031Rebuild the morning bookable-crew board100101/10197 (14 failed)9 minresolved
task-0032Build the weekly capability index100116/118114 (13 failed)10 minresolved
task-0033Keep the RecruitOS pipeline and the Placemint ledger in step0152/153110 (9 failed)12 minpartial
task-0034Weekly crew top-up is putting people on the gate twice100165/165123 (14 failed)12 minresolved
task-0035Sync Vettly subjects, checks, and reports into our canonical store022/2686 (13 failed)16 minmandatory fail
task-0036Harden the Placemint book extract for the larger tenant0209/420126 (16 failed)17 minpartial
task-0037Fix the Placemint redeployment sync100306/306123 (15 failed)13 minresolved
task-0038TalentLoop connector: make webhook-driven freshness durable against a misbehaving receiver018/37111 (11 failed)15 minmandatory fail
task-0039Build the nightly payroll bridge for the punch export10094/94101 (12 failed)12 minresolved
task-0040Push interview reschedules upstream into Interviewly014/17108 (12 failed)12 minmandatory fail
task-0041Paygrade connector: employee/assignment sync and writeback031/38112 (12 failed)13 minmandatory fail
task-0042Move the rota mirror onto instant storage100117/117113 (16 failed)9 minresolved
task-0043The rota warehouse is counting the same night more than once083/89106 (16 failed)14 minpartial
task-0044Nightly roster census disagrees with CrewCall about who works here10067/6787 (8 failed)9 minresolved
task-0045Replatform the ATS integration: StaffLine to Bullpen v2021/25130 (14 failed)12 minmandatory fail
task-0046Get the scorecard export into TalentLoop10089/8998 (12 failed)12 minresolved
task-0047Reconcile and safely apply the mobility stage repair queue100100/10076 (13 failed)13 minresolved
task-0048Migrate the Brackett closure archive onto Paygrade100211/211132 (18 failed)11 minresolved
task-0049Placemint connector: webhooks, polling, and writeback for a high-volume tenant10068/68106 (11 failed)15 minresolved
task-0050Fix the Placemint fee-corrections connector100254/254114 (14 failed)13 minresolved