Skip to content

fix(components): make per-model option scanning opt-in for Provider probes - #2990

Merged
lodyai[bot] merged 2 commits into
mainfrom
investigate-provider-config-refresh
Nov 20, 2026
Merged

lodyai[bot] merged 2 commits into
mainfrom
investigate-provider-config-refresh

Conversation

@lodyai

@lodyai lodyai Bot commented Nov 20, 2026 •

Copy link
Copy Markdown
Contributor

Summary

  • Make readModelOptions opt-in (default false) in useRefreshMachineAcpCapabilities, so Provider creation, testing, retrying in AgentConfigDialog and Lody Agent connection discovery run only session/new and do not block modals on serial per-model session/set_config_option sweeps (such as Devin's 169-model catalog).
  • Pass readModelOptions: true only from explicit Provider refreshes in Settings > Agents (ProviderRow), keeping inline per-model option discovery and progress reporting intact.
  • Remove unused per-model progress state/props from AgentConfigDialog.
  • Add a regression test for useRefreshMachineAcpCapabilities and update the EN/ZH architecture note (2026-11-11-observed-model-config-options).

Test plan

  • NODE_ENV=test pnpm --filter @lody/components test tests/provider-row-reauthentication.test.tsx tests/agent-config-dialog.test.tsx tests/lody-agent-setting.test.tsx (59/59 passed)
  • pnpm --filter @lody/components run typecheck
  • pnpm lint & pnpm exec oxfmt --check on all modified files
  • node scripts/check-i18n.mjs, check-code-collab-imports, check-platform-boundaries, node scripts/docs/main.mjs check

Generated with Devin

…robes

Creating, testing, or retrying a Provider in AgentConfigDialog (and
discovering Lody Agent connections) previously requested
`readModelOptions: true` unconditionally via
`useRefreshMachineAcpCapabilities`. For Providers with large model
catalogs such as Devin (169 models), this blocked the modal on serial
`session/set_config_option` switches for every model.

Make `readModelOptions` opt-in (defaulting to false) in
`useRefreshMachineAcpCapabilities`, pass `readModelOptions: true` only
from explicit Provider refreshes in Settings > Agents (`ProviderRow`),
remove the unused per-model progress state from `AgentConfigDialog`,
update the architecture note, and add a regression test.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Model: gemini-4-neon-medium
…s test

`activeWorkspaceRuntimeAtom` is derived from `runtimeAtom` and
`currentWorkspaceIdAtom`, so seed `runtimeAtom` with matching
`workspaceId` in the Jotai store for the regression test.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Model: gemini-4-neon-medium
@lodyai
lodyai Bot merged commit 57f72f3 into main Nov 20, 2026
13 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant