fix: add display 'summarized' for Opus 4.7 on Bedrock to prevent reasoning errors

[Dulani]

Claude Opus 4.7 on AWS Bedrock fails with "reasoning part 0 not found" error on first message when using extended thinking. Opus 4.7 defaults thinking.display to "omitted" instead of "summarized".

Steps to Reproduce

1. Configure OpenCode to use us.anthropic.claude-opus-4-20250514 on AWS Bedrock
2. Enable extended thinking (any variant)
3. Send a message

Expected vs Actual

• Expected: First message completes
• Actual: "reasoning part 0 not found" error

Why This Happens

Commit #22873 fixed this for Anthropic Messages API but not for Bedrock. When display: "omitted", thinking blocks arrive empty, breaking Vercel AI SDK's streaming transform.

Environment

• OpenCode 1.4.7+
• Provider: AWS Bedrock
• Model: us.anthropic.claude-opus-4-20250514 (Opus 4.7)

---

Authored by: Haiku 4.5 (90%), Human review & direction (10%)

[github-actions]

This issue doesn't fully meet our contributing guidelines.

What needs to be fixed:

• The issue footer explicitly states it was "Authored by: Haiku 4.5 (90%)" — AI-generated content is not acceptable per our contributing guidelines. Please rewrite this in your own words, describing what you actually observed and experienced.
• The title uses a PR-style prefix ("fix:") rather than following the Bug Report template format.

Please edit this issue to address the above within 2 hours, or it will be automatically closed.

---

This issue might be a duplicate of existing issues. Please check:

• fix: reasoning text trimEnd() in processor.ts invalidates Bedrock extended thinking signature #18078: Similar Bedrock extended thinking signature/reasoning block issue causing validation errors on multi-turn conversations
• bug(provider): interleaved reasoning filter produces empty content arrays, crashing Bedrock sessions #17705: Related Bedrock crash due to reasoning block handling producing invalid content arrays
• normalizeMessages() removes empty text parts between reasoning blocks, invalidating Anthropic thinking block signatures #16748: normalizeMessages() removes empty text parts between reasoning blocks, invalidating Anthropic thinking block signatures

If you believe this was flagged incorrectly, please let a maintainer know.

[github-actions]

This issue has been automatically closed because it was not updated to meet our contributing guidelines within the 2-hour window.

Feel free to open a new issue that follows our issue templates.