{"data":{"slug":"ruc-datalab-coda-bench","name":"CoDA-Bench","tagline":"Benchmark for code agents on data-intensive tasks","github_url":"https://github.com/ruc-datalab/CoDA-Bench","owner":"ruc-datalab","repo":"CoDA-Bench","owner_avatar_url":"https://avatars.githubusercontent.com/u/76154266?v=4","primary_language":"Python","stars":45,"forks":1,"topics":["agent","agentic","agentic-ai","ai","benchmark","code-agent","code-generation","coding-agent","data-analysis","data-engineering","data-intensive","data-science","evaluation","icml","kaggle","llm","llm-agent","llm-evaluation","sandbox"],"archived":false,"github_pushed_at":"2026-06-17T01:34:24+00:00","maintenance_label":"Steady","stars_delta_30d":3,"url":"https://www.graphcanon.com/tools/ruc-datalab-coda-bench","markdown_url":"https://www.graphcanon.com/tools/ruc-datalab-coda-bench.md","api_url":"https://www.graphcanon.com/api/graphcanon/tools/ruc-datalab-coda-bench","graph_url":"https://www.graphcanon.com/api/graphcanon/graph?tool=ruc-datalab-coda-bench","description":"CoDA-Bench is a benchmark for code agents on data-intensive tasks. 🎈代码智能体能搞定数据密集型任务吗? ","homepage_url":"https://coda-bench.github.io/","license":"MIT","open_issues":3,"watchers":1,"ai_summary":"CoDA-Bench is designed to evaluate the performance of AI-powered code agents in handling complex data-related tasks.","readme_excerpt":"## 🚀 Quick Start\n\n<img src=\"./assets/overview.png\" style=\"height: 20em\">\n\n---\n\n### Installation\n\n```bash\ngit clone https://github.com/ruc-datalab/CoDA-Bench.git\ncd CoDA-Bench\npip install -e .\n```\n\n---\n\n### Run Evaluation (Docker Mode)\n\n**Step 1: Build Docker Image**\n```bash\ncd docker\n./build_all.sh\ncd ..\n```\n\n**Step 2: Set API Credentials**\n```bash\nexport LLM_API_KEY=\"your-api-key\"\nexport LLM_BASE_URL=\"https://api.openai.com/v1\"  # Optional\n```\n\n**Step 3: Run Evaluation**\n```bash\n\n---\n\n### Why Docker?\n\nDocker mode provides **secure isolation**:\n- ✅ Agents cannot access benchmark answers\n- ✅ Network-restricted environment (only LLM API accessible)\n- ✅ Resource limits (memory, CPU, timeout)\n- ✅ Reproducible across different machines\n\nSee [docker/README.md](docker/README.md) for detailed Docker documentation.\n\n---\n\n## 📄 License\n\nMIT License. See [LICENSE](LICENSE) for details.\n\nIndividual Kaggle datasets may have their own licenses.","github_created_at":"2026-05-29T04:39:31+00:00","created_at":"2026-07-15T10:39:43.675896+00:00","updated_at":"2026-09-20T04:24:58.081385+00:00","categories":[{"slug":"ai-agents","name":"AI Agents","url":"https://www.graphcanon.com/categories/ai-agents","markdown_url":"https://www.graphcanon.com/categories/ai-agents.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/ai-agents"},{"slug":"evaluation-observability","name":"Evaluation & Observability","url":"https://www.graphcanon.com/categories/evaluation-observability","markdown_url":"https://www.graphcanon.com/categories/evaluation-observability.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/evaluation-observability"}],"tags":[{"slug":"agent","name":"agent"},{"slug":"agentic-ai","name":"agentic-ai"},{"slug":"benchmark","name":"benchmark"},{"slug":"code-agent","name":"code-agent"},{"slug":"data-engineering","name":"data-engineering"},{"slug":"icml","name":"icml"},{"slug":"llm-evaluation","name":"llm-evaluation"}],"trust":{"provenance":{"is_fork":false,"github_id":1253011729,"owner_type":"Organization","methodology":"github_public_v1","parent_repo":null,"near_duplicate_slugs":[]},"computed_at":"2026-09-09T06:00:29.348Z","maintenance":{"label":"Steady","score":60,"methodology":"github_public_v1","releases_90d":0,"days_since_push":84,"last_release_at":null,"stars_delta_30d":3,"open_issues_delta_30d":1},"security_summary":{"status":"no_lockfile","scanner":null,"low_count":0,"high_count":0,"last_scan_at":"2026-07-15T10:39:45.103Z","medium_count":0,"scan_profile":"none","critical_count":0}},"capability_facts":{"scan":{"source":"repo_scan","observed_at":"2026-09-09T06:00:29.905Z"},"has_cli":{"value":true,"source":"pyproject.toml:[project.scripts]","observed_at":"2026-09-09T06:00:29.905Z"},"languages":{"value":["python"],"source":"github.language+pyproject.toml","observed_at":"2026-09-09T06:00:29.905Z"},"license_spdx":{"value":"MIT","source":"github.license","observed_at":"2026-09-09T06:00:29.905Z"}},"decision_facts":{"hosting":null,"pricing":{"model":"freemium","summary":"Available under the MIT License, free to use but may require additional costs for API credentials if using external services like OpenAI's APIs."},"requirements":{"min_ram_gb":null,"requires_docker":true},"constraints":{"min_ram_gb":null,"pricing_model":"freemium","requires_docker":true},"when_to_use":["When you aim to evaluate the reliability and security of an AI-powered code agent with strict control over network access and data isolation.","For benchmarking purposes specifically involving complex, data-related tasks where reproducibility across different machines is crucial."],"when_not_to_use":["If your project does not require Docker-level secure isolation or if the overhead of setting up a Docker environment is prohibitive for your workflow.","When you are dealing with less complex data tasks that do not demand stringent security measures such as restricted network environments and resource limits."],"source":"enrich:decision_facts","observed_at":"2026-07-16T18:46:04.010Z"},"constraint_facets":{"min_ram_gb":null,"pricing_model":"freemium","requires_docker":true},"decision_summary":[{"label":"Pricing","value":"freemium - Available under the MIT License, free to use but may require additional costs for API credentials if using external services like OpenAI's APIs."},{"label":"Requirements","value":"Requires Docker"},{"label":"Adopt for","value":"CoDA-Bench provides secure isolation for evaluating code agents in data-intensive tasks via Docker-mode execution."}]}}