[Ready for Review] Adapter | Condense tutorial for human, create agent version separately

[crystalxyz]

Summary

This PR was motivated by a prior Harbor meeting discussion saying that the current harbor adapter tutorial is really long and hard to follow for human readers. In order to lower the entrance boundary and make it easier for people to contribute, we decide to condense the adapter tutorial for human readers so that they can easily keep track of their progress. Also, a separate agent-versioned tutorial is created so that contributors can easily use agents to build adapters.

New structure

• /doc/datasets/adapter (Title: Adapters) -> Human version tutorial
• /doc/datasets/adapter-ai (Title: Adapters (Agent Version)) -> Agent version tutorial
• A block added at the beginning of the human tutorial to point agents to the other page

Details

• The original adapter.mdx file was renamed to adapter-ai.mdx, and this file was modified to work better with agents. The key modifications include replacing the accordion UI with tables and lists; providing detailed field type and explanation for each json file; defining clear steps and validation criteria
• A new adapter.mdx is created from the original file, but as a concise human-readable version

[Ternura143]

LGTM! I only have a small issue that needs to be addressed.

[Ternura143]

parity_sampling_rate formula is inverted in the schema — should be parity_benchmark_size / adapted_benchmark_size, not adapted_benchmark_size / parity_benchmark_size.

[AlienKevin]

For cases where the original benchmark has broken oracles, maybe we can advice the agent to document the tasks with oracle issues and file bugs to the upstream benchmark instead of attempting to fix them on the Harbor side.