Community Data License Agreement – Permissive – Version 2.0

This is the Community Data License Agreement – Permissive, Version 2.0 ("Agreement"). 
Data Provider and Data Recipient each agree as follows:

1. Provision of the Data

   1.1. A Data Provider may provide Data to Data Recipients under this Agreement.

   1.2. A Data Recipient may receive Data under this Agreement.

2. License Grant

   2.1. Subject to the terms and conditions of this Agreement, Data Provider grants to 
        Data Recipient a worldwide, non-exclusive, royalty-free, perpetual (subject to 
        termination under Section 5) license to:

        (a) use, reproduce, extract, copy, modify, and create derivative works of the Data;
        (b) publicly perform, publicly display, and distribute the Data and derivative works; and
        (c) sublicense rights in the Data and derivative works.

3. Data Provider Obligations

   3.1. Data Provider will make available to Data Recipient the Data in a machine-readable 
        format under the terms of this Agreement.

4. Data Recipient Obligations

   4.1. Data Recipient will comply with all applicable laws and regulations in its use of 
        the Data.

5. Termination

   5.1. This Agreement will terminate automatically if Data Recipient materially breaches 
        the terms of this Agreement.

6. Disclaimer of Warranties and Limitation of Liability

   6.1. Except as expressly set forth in this Agreement, the Data is provided on an "AS IS" 
        basis, without warranties or conditions of any kind, either express or implied.

   6.2. Neither Data Provider nor Data Recipient will be liable for any damages arising out 
        of or related to this Agreement.

7. Miscellaneous

   7.1. This Agreement constitutes the entire agreement between Data Provider and Data 
        Recipient with respect to the Data.

Full license text available at: https://cdla.dev/permissive-2-0/

================================================================================

DATA ATTRIBUTION

Source: Open Agent Leaderboard
Website: https://exgentic.ai
GitHub: https://github.com/Exgentic/open-agent-leaderboard
Paper: https://arxiv.org/abs/2602.22953

Description:
This dataset contains benchmark results for general-purpose AI agent systems
evaluated across six diverse tasks: AppWorld, BrowseComp+, SWE-bench, and
TauBench (Airline, Retail, Telecom). The data includes success rates, costs,
and performance metrics for different agent-model combinations.