Conversation
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
|
Warning Review limit reachedNext included review available in 11 minutes. View limit detailsLimit details: You’ve used the included review currently available. You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Review configuration: ⚙️ Run configurationConfiguration used: Repository: jmrplens/gitlab-mcp-server/.coderabbit.yaml Review profile: CHILL Plan: Advanced Run ID: 📒 Files selected for processing (4)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
jmrplens
added this pull request to stack #863
September 20, 2026 10:16
github-actions
Bot
added
bug
jmrplens
removed this pull request from stack #863
September 20, 2026 19:40
jmrplens
added this pull request to stack #866
September 20, 2026 19:40
All six release-link descriptions, the doc comment above them and the release reference page named an `external` field beside the five the Output type carries. GitLab has not sent it since 16.0: the live oracle taken from 19.3.1 records API::Entities::Releases::Link as exposing id, name, url, direct_asset_url and link_type, with no conditions and nothing else. A model was told six times to expect a key it never receives, and the likely answer to that is a second call looking for the field or a value invented in its place. The word goes rather than the field arriving. client-go still models ReleaseLink.External, so filling an output field from it is the tempting reading and the wrong one: it would publish a bool that is false for every link on every instance, which is the class of defect issue 580 was about. That stale SDK field is already recorded in docs/development/upstream-bugs.md, among the ten modelled fields no Grape entity exposes, so nothing new is owed upstream. Sixteen mock responses in the package fed `external` back and a helper set the SDK field from a parameter of its own, so every fixture here described a link no instance can serve, which is what let the promise survive a green suite for as long as it did. They go with it, and a guard takes their place: the six descriptions are held against the JSON keys Output publishes, with the candidates taken from client-go's own struct rather than from a list naming `external`, so the next key the SDK models and this server does not is caught the moment a description promises it. The batch action also recovers the guidance a deleted duplicate block carried. It told a model not to call link_create once per asset and not to put a single name and url at top level, and a later assignment overwrote the whole block a few lines down, so none of it ever reached a model. Its usage sentence and both confusions are in the live block now, and release.link_create is among its related actions, where the dead copy had it and the live one did not. Generated artifacts move and are left for the regeneration at the top of the stack: testdata/tools_individual.json (six tools), tools_meta.json, tools_meta_compact.json and tools_meta_full.json (one each), the llms.txt family and the testing reference, which gains one test function. (cherry picked from commit 91aa4b0f2244e5533dbfe8be9aa37094a4cd3101)
|
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
1 participant
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.
All six release-link descriptions, the doc comment above them and the release reference page named an
externalfield beside the five the Output type carries. GitLab has not sent it since 16.0: the live oracle taken from 19.3.1 records API::Entities::Releases::Link as exposing id, name, url, direct_asset_url and link_type, with no conditions and nothing else. A model was told six times to expect a key it never receives, and the likely answer to that is a second call looking for the field or a value invented in its place.The word goes rather than the field arriving. client-go still models ReleaseLink.External, so filling an output field from it is the tempting reading and the wrong one: it would publish a bool that is false for every link on every instance, which is the class of defect issue 580 was about. That stale SDK field is already recorded in docs/development/upstream-bugs.md, among the ten modelled fields no Grape entity exposes, so nothing new is owed upstream.
Sixteen mock responses in the package fed
externalback and a helper set the SDK field from a parameter of its own, so every fixture here described a link no instance can serve, which is what let the promise survive a green suite for as long as it did. They go with it, and a guard takes their place: the six descriptions are held against the JSON keys Output publishes, with the candidates taken from client-go's own struct rather than from a list namingexternal, so the next key the SDK models and this server does not is caught the moment a description promises it.The batch action also recovers the guidance a deleted duplicate block carried. It told a model not to call link_create once per asset and not to put a single name and url at top level, and a later assignment overwrote the whole block a few lines down, so none of it ever reached a model. Its usage sentence and both confusions are in the live block now, and release.link_create is among its related actions, where the dead copy had it and the live one did not.
Generated artifacts move and are left for the regeneration at the top of the stack: testdata/tools_individual.json (six tools), tools_meta.json, tools_meta_compact.json and tools_meta_full.json (one each), the llms.txt family and the testing reference, which gains one test function.
(cherry picked from commit 91aa4b0f2244e5533dbfe8be9aa37094a4cd3101)