public benchmark leaderboards

Ranked benchmark results grouped by task, runtime, risk profile, and execution backend, with derived risk and environment aliases for easier browsing.

Groups: 1 · Packages: 1 · Runs: 1

Generated: 2026-03-16T17:29:46.327Z

Filters: benchmarkId=benchmark/research-brief-orchestration@1.0.0

leaderboard JSON · snapshot payload.json

benchmark/deploy-change-audit@1.0.0 benchmark/docs-migration-plan@1.0.0 benchmark/policy-safety-review@1.0.0 benchmark/release-notes-synthesis@1.0.0 benchmark/release-review-orchestration@1.0.0 benchmark/research-brief-orchestration@1.0.0 benchmark/source-backed-research@1.0.0 benchmark/treasury-briefing@1.0.0 runtime: bun runtime: claude runtime: codex risk: high risk: low risk: medium env: local

Research brief orchestration

Benchmark: benchmark/research-brief-orchestration@1.0.0

Runtime: codex · Risk: low · Risk profile: default:none · Env: local · Backend: local

Sandbox profile: default · Network policy: none

Runs: 1 · Successes: 1

#1 web/recipes-live.example/research-brief-recipe@1.0.427100 · avg 100.0% · best 100.0% · runs 1 · successes 1 · latest run