{"schemaVersion":"agentdepot.eval_plan.v1","asOf":"2026-10-28","storesIpAddresses":false,"query":"browser","matchMode":"matched_task","taskType":"browser","label":"Web Browser, DOM Navigation & Scraping/Action Agents","recommendedBenchmarks":[{"id":"webarena-browsergym","name":"WebArena & BrowserGym","officialUrl":"https://webarena.dev/","graderType":"DOM state, URL & DB post-condition check"}],"depotBenchmarkSuite":{"type":"browser_web","tasksApi":"/api/benchmark/tasks?type=browser_web","submitApi":"/api/benchmark/runs"},"stackPicks":{"evalHarness":{"jtbdId":"computer-use-browser-agent-evals","pickProviderId":"browserbase","coPickProviderId":"e2b","runnersUpIds":["e2b","steel-dev"]},"sandboxOrRunner":{"jtbdId":"cloud-browser","pickProviderId":"browserbase","coPickProviderId":"steel-dev","runnersUpIds":["steel-dev","browserless"]},"ciGate":{"jtbdId":"agent-trajectory-tool-use-evals","pickProviderId":"braintrust","coPickProviderId":"langsmith","runnersUpIds":["langsmith","arize-phoenix"]},"judgeCalibration":{"jtbdId":"cost-latency-per-task-benchmarking","pickProviderId":"langfuse","coPickProviderId":"portkey","runnersUpIds":["portkey","openrouter"]}},"targetMetrics":["DOM/DB post-condition pass@1","pass^3 across clean browser contexts","steps-per-task","browser-hours + token USD per solved task"],"antiContaminationRule":"Reset web app state and browser cookies before every episode so prior trajectories never leak state.","nextCalls":["/api/decide?jtbd=computer-use-browser-agent-evals","/api/benchmark/tasks?type=browser_web","/api/playbooks/agent-evals"]}