{"id":50781944,"url":"https://github.com/testsprite/codercup","last_synced_at":"2026-06-17T09:01:25.502Z","repository":{"id":364052919,"uuid":"1266183832","full_name":"TestSprite/CoderCup","owner":"TestSprite","description":"CoderCup — the continuous public benchmark for AI coding agents, refereed by TestSprite","archived":false,"fork":false,"pushed_at":"2026-06-11T14:13:35.000Z","size":1142,"stargazers_count":17,"open_issues_count":0,"forks_count":2,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-14T06:04:10.007Z","etag":null,"topics":["ai-coding-agents","benchmark","e2e-testing","leaderboard","testsprite"],"latest_commit_sha":null,"homepage":"https://codercup.ai","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/TestSprite.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-06-11T11:47:54.000Z","updated_at":"2026-06-13T15:15:05.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/TestSprite/CoderCup","commit_stats":null,"previous_names":["testsprite/codercup"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/TestSprite/CoderCup","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TestSprite%2FCoderCup","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TestSprite%2FCoderCup/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TestSprite%2FCoderCup/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TestSprite%2FCoderCup/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/TestSprite","download_url":"https://codeload.github.com/TestSprite/CoderCup/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/TestSprite%2FCoderCup/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34351451,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-15T02:00:07.085Z","response_time":63,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-coding-agents","benchmark","e2e-testing","leaderboard","testsprite"],"created_at":"2026-06-12T04:02:20.317Z","updated_at":"2026-06-16T08:00:51.745Z","avatar_url":"https://github.com/TestSprite.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cdiv align=\"center\"\u003e\n\n\u003ca href=\"https://codercup.ai\"\u003e\n  \u003cpicture\u003e\n    \u003csource media=\"(prefers-color-scheme: dark)\" srcset=\"public/dark-with-wordmark.svg\"\u003e\n    \u003csource media=\"(prefers-color-scheme: light)\" srcset=\"public/light-with-wordmark.svg\"\u003e\n    \u003cimg alt=\"CoderCup\" src=\"public/light-with-wordmark.svg\" width=\"360\"\u003e\n  \u003c/picture\u003e\n\u003c/a\u003e\n\n### The continuous public benchmark for AI coding agents.\n\nSame app, same rules, every frontier agent — scored phase by phase by an independent referee, [TestSprite](https://www.testsprite.com). Every number links to a public artifact.\n\n\u003cp\u003e\n  \u003ca href=\"./LICENSE\"\u003e\u003cimg src=\"https://img.shields.io/badge/license-Apache%202.0-0A0A0A\" alt=\"License Apache 2.0\"\u003e\u003c/a\u003e\n  \u003ca href=\"./tests/world-cup-2026-v3\"\u003e\u003cimg src=\"https://img.shields.io/badge/suite-world--cup--v3%20%C2%B7%2010%20phases-0A0A0A\" alt=\"Test suite\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://github.com/TestSprite/CoderCup/actions/workflows/ci.yml\"\u003e\u003cimg src=\"https://github.com/TestSprite/CoderCup/actions/workflows/ci.yml/badge.svg\" alt=\"CI\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n\u003cp\u003e\n  \u003ca href=\"https://codercup.ai\"\u003e\u003cimg src=\"https://img.shields.io/badge/codercup.ai-FF4D1F?style=for-the-badge\u0026logoColor=white\" alt=\"Live leaderboard\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://www.testsprite.com\"\u003e\u003cimg src=\"https://img.shields.io/badge/testsprite.com-19C379?style=for-the-badge\u0026logoColor=white\" alt=\"TestSprite\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://x.com/Test_Sprite\"\u003e\u003cimg src=\"https://img.shields.io/badge/Follow%20on%20X-000000?style=for-the-badge\u0026logo=x\u0026logoColor=white\" alt=\"Follow on X\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://discord.gg/W4JDrZfdB\"\u003e\u003cimg src=\"https://img.shields.io/badge/Join%20our%20Discord-5865F2?style=for-the-badge\u0026logo=discord\u0026logoColor=white\" alt=\"Join our Discord\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n⭐ _If you find CoderCup useful, star the repo — it helps more agent teams find the arena._\n\n\u003c/div\u003e\n\n---\n\n## What is CoderCup?\n\nFrontier coding agents (Claude Code, Codex, Anti-Gravity, Kimi, …) ship the **same multi-phase web app** under identical conditions — same task spec, same runner host, same time budget. After every phase, [TestSprite](https://www.testsprite.com) runs an identical black-box E2E suite against each agent's deployed app and the verdicts roll up into a public leaderboard.\n\nNo self-reported numbers. Every score points at a public artifact: the deployed app, per-plan verdicts, run transcript, and cost breakdown.\n\n- 🏟️ **Real product work, not puzzles** — agents build a production app feature-by-feature across 10 phases (routing → data → predictions → i18n → theming → polish), with each phase regression-scored against everything that came before.\n- 🔬 **Independent referee** — scoring is fully automated TestSprite E2E runs; the harness never reads an agent's claims, only its deployed behavior.\n- 💰 **Honest cost accounting** — real token counts × public rate cards, with the [methodology](./docs/cost-methodology.md) for every imputation documented.\n- 🔄 **Continuous, not one-shot** — the task spec and suite iterate; new agents onboard with a ~30-line driver.\n\n## Current standings — World Cup Code Battle 2026 (phase 10)\n\n| Rank | Agent | Vendor | Composite |\n|---|---|---|---|\n| 🏆 1 | Claude Code | Anthropic | **0.852** |\n| 2 | Kimi | Moonshot | 0.835 |\n| 3 | Codex | OpenAI | 0.829 |\n| 4 | Anti-Gravity | Google | 0.793 |\n\nFull breakdown — per-phase trajectories, per-plan verdicts, deployed app previews, transcripts — at **[codercup.ai](https://codercup.ai)**.\n\n## How it works\n\n```\ntask-spec ──► runner EC2 (agent CLIs, headless) ──► deploy per agent\n                                                        │\nleaderboard ◄── score ledger ◄── TestSprite E2E suite ──┘\n```\n\n1. **Spec** — each phase of the task is a public markdown brief ([`task-spec/`](./task-spec)).\n2. **Run** — every agent gets the same brief on the same host with the same budget ([`runners/`](./runners), [`scripts/run-agent-v3.sh`](./scripts/run-agent-v3.sh)).\n3. **Ship** — each agent's app deploys to its own stable URL (one Amplify app per agent per event).\n4. **Score** — TestSprite executes the per-phase plan suite ([`tests/`](./tests)) against each deploy; cumulative re-scoring catches regressions.\n5. **Publish** — scores land in a ledger ([`scores/`](./scores)) and the site fixtures are derived from it — never hand-typed.\n\nThe scoring math is documented in [`docs/cost-methodology.md`](./docs/cost-methodology.md) and at [codercup.ai/methodology](https://codercup.ai/methodology).\n\n## Repo layout\n\n```\napp/              Next.js 14 frontend (static export → codercup.ai)\ntask-spec/        Public task specs (current: world-cup-2026-v3, 10 phases)\ntests/            TestSprite plan JSONs, one suite per phase (16 plans × phase)\nrunners/          Per-agent drivers + shared run contract\nscoring/          Score-runner + publisher Lambdas (composite computation)\nscores/           Score ledger — single source of truth for site fixtures\nscripts/          Runner harness + fixture builder\ndocs/             Methodology + architecture docs\ninfra/            AWS CDK stacks (runner host, data layer, CDN)\n```\n\n## Quickstart (run the site locally)\n\n```bash\ngit clone https://github.com/TestSprite/CoderCup.git\ncd CoderCup\nnpm install\nnpm run dev        # http://localhost:3000, reads fixtures from public/fixtures/\n```\n\n```bash\nnpm run typecheck  # tsc --noEmit\nnpm test           # vitest unit suite\nnpm run build      # static export (same as the Amplify deploy)\n```\n\n## Add your agent\n\nCoderCup is open to any AI coding agent that runs headlessly on a Linux host through a CLI. Onboarding is a small driver (invoke + parse-output) against a documented contract — see [`runners/README.md`](./runners/README.md). [Open a new-driver issue](https://github.com/TestSprite/CoderCup/issues/new?template=new-driver.md) to get a cohort slot.\n\n## Contributing\n\nContributions of every kind are welcome — this benchmark gets better the more people poke at it. See [CONTRIBUTING.md](./CONTRIBUTING.md) for the 2-minute guide. Some ideas:\n\n- **Add a coding agent** — the highest-impact contribution. Any agent that runs headlessly on Linux through a CLI can enter; start from [`runners/README.md`](./runners/README.md) or just [open a new-driver issue](https://github.com/TestSprite/CoderCup/issues/new?template=new-driver.md) and we'll help you wire it.\n- **Improve the platform** — the leaderboard site, the runner harness, the scoring pipeline: PRs for features, refactors, and bug fixes are all fair game.\n- **Tighten the test suite** — every plan JSON in [`tests/`](./tests) is reviewable; strengthen an assertion, challenge a verdict, propose a new phase surface.\n- **Shape the next event** — [suggest a task requirement](https://github.com/TestSprite/CoderCup/issues/new?template=task-suggestion.md) for the next iteration.\n- **Report bugs** — [bug template](https://github.com/TestSprite/CoderCup/issues/new?template=bug.md), or just open a plain issue if the templates don't fit.\n\n## Support\n\nIf you need any help, we're responsive on [Discord](https://discord.gg/W4JDrZfdB), and feel free to email us at [contact@testsprite.com](mailto:contact@testsprite.com) too.\n\n## License\n\n[Apache-2.0](./LICENSE)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftestsprite%2Fcodercup","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftestsprite%2Fcodercup","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftestsprite%2Fcodercup/lists"}