feat: Add Scitely AI as built-in provider

[YlunoZup]

Closes #9429

• Add Scitely AI provider with 15 models including
  Qwen3, DeepSeek, Kimi, and GLM

  • Add Scitely to PROVIDER_PRIORITY in dialog for
    visibility
  • Add Scitely documentation in providers.mdx

  Models included:

  1. tstars2.0
  2. qwen3-coder-plus
  3. qwen3-max
  4. qwen3-vl-plus
  5. qwen3-max-preview ✨ (added)
  6. kimi-k2-0905 ✨ (added)
  7. glm-4.6
  8. kimi-k2
  9. deepseek-v3.2
  10. deepseek-r1
  11. deepseek-v3
  12. qwen3-32b
  13. qwen3-235b-a22b-thinking-2507
  14. qwen3-235b-a22b-instruct
  15. qwen3-235b

  Scitely provides FREE API access to these models
  for our partners and community at
  https://api.scitely.com/v1

What does this PR do?

Adds Scitely AI as a built-in provider for OpenCode,
giving users FREE access to
powerful AI models without manual JSON configuration.

Scitely offers free API access to models like Qwen3,
DeepSeek R1, Kimi K2, and more
for our partners and community.

Changes:

• Added Scitely provider definition with 14 models in
  provider.ts
• Added Scitely to the Popular providers category in the
  provider dialog
• Added documentation for Scitely setup in
  providers.mdx

Users can now:

1. Run /connect and select "Scitely"
2. Enter their API key (free to obtain)
3. Access all Scitely models via /models

How did you verify your code works?

• Verified TypeScript types match the existing provider
  patterns
  • Followed the same structure as other OpenAI-compatible
    providers
  • Models are configured with correct context limits and
    capabilities
  • Documentation follows the same format as other
    provider docs

[github-actions]

Thanks for your contribution!

This PR doesn't have a linked issue. All PRs must reference an existing issue.

Please:

1. Open an issue describing the bug/feature (if one doesn't exist)
2. Add Fixes #<number> or Closes #<number> to this PR description

See CONTRIBUTING.md for details.

[github-actions]

The following comment was made by an LLM, it may be inaccurate:

No duplicate PRs found

[neriousy]

I don't know if the team wants to add a built-in 3rd party provider. If people would like to use your API they can just paste it in the config. No need to hardcode it in, especially when you're hardcoding models which will probably change in the future.
You would need to send another PR to adjust that.