release: v7.13.0 — LLM-driven GUI personalization - #580
Merged
Conversation
… cleanup Addresses review feedback on #577 (organization + correctness): - Extract the ~150-line /personalize-gui-page endpoint body into a new services/utils/gui_personalization_utils.py, mirroring the agent_personalization pattern: the ~40-line LLM system prompt is now a module constant, and personalize_gui_page() owns validation, prompt building, the OpenAI call and response parsing. The router endpoint is now a thin validate -> delegate -> shape-response handler. - Consolidate the OpenAI model default into one DEFAULT_OPENAI_MODEL = 'gpt-5' constant. Removes the inconsistent/invalid defaults the PR introduced (gpt-5.4-mini x7, gpt-5.6-luna) that would 404 at runtime. - Remove a leftover print('calling now') debug statement. - Remove two dead '... and False' branches in replace_reply_batch. - Add test_gui_personalization.py (10 tests, LLM mocked) covering validation, model selection, response parsing and error mapping. Left as-is: the translate-before-style reorder in replace_reply_batch (appears intentional); pre-existing gpt-5.5 default at generation_router line 172 (out of scope).
…anup refactor(gui-personalization): extract endpoint to a module + fix model defaults
Feature/gui personalization
…ion skip Regression tests for the #579 fixes in object_diagram_processor: - a manually-typed 'title == X' criterion must split on the stored '==' (not the first '='), so the value is 'X' not '= X'; - a blank criterion ('title = ') is skipped, not materialised as an empty-valued slot.
fix user profile ocl errors
- Bump setup.cfg 7.12.0 -> 7.13.0 and add v7.13.0 release notes. - Frontend submodule bump 360dcfa6 -> b14dadec: the GUI-personalization UI (user-profile form editor, per-profile page variants + switching, web-app version selection in the generation dialog) plus review cleanups. All GUI-personalization work; no multilingual changes. - Backend: the GUI-personalization feature (#577), the extract-to-service cleanup (#578), and the user-profile OCL/parsing fixes + regression tests (#579) are already on development.
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
2 participants
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.
Summary
POST /personalize-gui-pageendpoint LLM-adapts a GrapesJS page ({components, css}) to a user-profile model; logic lives in a dedicatedgui_personalization_utilsservice (thin router).UserDiagrammodels.age == 22splits on the stored==, blank criteria skipped), OCL evaluation skips constraints on unset attributes, and backend cleanup (singleDEFAULT_OPENAI_MODEL, dead code removed, regression tests).360dcfa6 → b14dadec(all GUI-personalization work; no multilingual changes).Test plan
python -m pytest tests/utilities/web_modeling_editor/backend— converter + gui-personalization suites greencd docs && make html— the new v7.13.0 page renders