fix(soul): name the config fix in the unsupported-capability error - #2590
Open
ayaangazali wants to merge 1 commit into
Open
fix(soul): name the config fix in the unsupported-capability error#2590ayaangazali wants to merge 1 commit into
ayaangazali wants to merge 1 commit into
Conversation
The error stated which capability was missing but not how to resolve it. For manually configured models, input capabilities like image_in come only from the model entry's `capabilities` setting, so a user hitting this had no indication of what to change. The message now names that setting and shows the exact value to declare. Two inline snapshots pinned the old text and were regenerated.
![devin-ai-integration[bot]](https://avatars.githubusercontent.com/in/811515?s=60&v=4){kind=small}
5 tasks
This was referenced Aug 6, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Related Issue
Partially addresses #2588 (the "no hint at the fix" half).
Description
#2588 reports two separate problems. This PR only tackles the smaller, self-contained one: the error tells you which capability is missing but never tells you what to change.
Before:
After:
The wording is deliberately conditional.
derive_model_capabilities(llm.py:532) only infers thinking-related capabilities, from the model name and a couple of hardcoded Kimi models, soimage_inandvideo_infor a custom model come solely from thecapabilitiesentry that is already documented indocs/en/configuration/config-files.md. The error should not claim the model is capable, only point at the setting if it is.What this PR does not do: the bigger half of #2588 is that the abort happens after tool side effects have landed, and the argument there is that a statically knowable capability should be checked before the run starts rather than when the offending message appears. That is a real design question about where the check belongs, so it seemed better left to you than guessed at by a drive-by.
Two existing inline snapshots pinned the old string (
tests/core/test_exceptions.py,tests_e2e/test_wire_errors.py); both were regenerated with--inline-snapshot=fixrather than hand-edited.Checklist
tests/core/test_llm_not_supported_message.pycovering the single and multiple capability wording; both fail on current main)make gen-changelogto update the changelog. (hand-edited CHANGELOG.md in the same style, I do not have the Kimi API setup the skill needs)make gen-docsto update the user documentation. (thecapabilitiessetting is already documented, nothing to regenerate)usual disclosure: freshman, so i deliberately took only the part of that issue i could be confident about and left the "check capabilities before running" design call to you. checked the capability resolution path myself and talked the wording through with claude code so the message would not overclaim. happy to reword or drop it :)