Repository navigation
vendor: resync deepcode-core to upstream 03c7ebb; drop merged reasoning-effort patch - #9
Merged
Merged
Conversation
reasoning_effort top-level fix was merged upstream (lessweb/deepcode-cli#370), so the local patch is dropped; the vendored engine also gains the upstream follow-up that type-casts 'max' effort in the edit tool's LLM helpers. markdown-it bump and README/vscode commits upstream do not touch the vendored core. Dependency versions unchanged.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Re-vendor
packages/corefrom lessweb/deepcode-cli@03c7ebbviascripts/sync-upstream.sh.reasoning-effort-top-level.patch— the fix was merged upstream in fix(core): send reasoning_effort at the top level of the request lessweb/deepcode-cli#370 (commits3ed02da+5954625). Behavior is identical; the e2e assertionreq.reasoning_effort === "high"passes unchanged.edit-handler.tscasts request params toChatCompletionCreateParamsNonStreamingso DeepSeek'smaxeffort type-checks in the edit tool's LLM helpers (old-string-not-found inference, escaped-string correction). No behavior change on our side, but unblocks futuremaxusage in those helpers.Upstream's other new commits (markdown-it bump, DeepCode Plus README section, vscode marketplace URLs) don't touch the vendored core; dependency versions in package.json match upstream core exactly.
Tests
npm test: 12/12 pass (build + e2e vs fake DeepSeek server, including the reasoning_effort and slow-pipe suites).