feat: Refactor and Unify the ModelInput Selector Across Components

[erichare]

Summary

This PR unifies model-selection behavior across Langflow components by moving shared logic into unified_models and updating the affected components to use it.

It introduces shared helpers for:

• refreshing/caching model options
• showing and hiding provider-specific fields
• pre-populating provider fields from stored variables
• constructing embedding clients from the same model metadata used by the UI

It also migrates knowledge ingestion onto the unified embedding path and preserves backward compatibility for older agent configuration styles.

What Changed

• Added get_embeddings() to unified_models to centralize embedding client construction.
• Added handle_model_input_update() to centralize ModelInput update behavior.
• Updated LanguageModelComponent, EmbeddingModelComponent, AgentComponent, ToolCallingAgentComponent, and other model-aware components to use the shared helper.
• Updated knowledge ingestion to store full embedding model selection metadata and rebuild embedding clients from that metadata.
• Restored KB creation support for ad hoc API keys during creation, not just globally configured credentials.
• Preserved legacy AgentComponent behavior for agent_llm / model_name style inputs.
• Refreshed related tests and generated assets/starter projects.

Why

Before this change, multiple components reimplemented similar logic for:

• model option refresh
• provider field visibility
• provider-specific credential fields
• embedding client instantiation

That duplication made behavior drift likely and increased maintenance cost. This PR consolidates that logic so model-aware components behave consistently and future provider changes can be made in one place.

User Impact

• Model dropdowns behave more consistently across components.
• Provider-specific fields such as WatsonX/Ollama settings are managed through shared logic.
• Knowledge ingestion can reconstruct embedding clients from saved metadata instead of relying on older string-only metadata.
• Creating a knowledge base still works when users provide an API key directly during setup.
• Existing agent setups using legacy configuration patterns continue to work.

[coderabbitai]

Important

Review skipped

Auto incremental reviews are disabled on this repository.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 47c8cda3-d4ac-43aa-9a63-6604afa9f219

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

Walkthrough

The PR refactors LanguageModelComponent, StructuredOutputComponent, and AgentComponent across 24 starter project JSON files, replacing in-method update_build_config logic with centralized delegation to handle_model_input_update. This consolidates configuration updates by removing bespoke provider-field handling and unifying option processing.

Changes

Cohort / File(s)	Summary
LanguageModelComponent Update Refactoring src/backend/base/langflow/initial_setup/starter_projects/Basic Prompt Chaining.json, Basic Prompting.json, Blog Writer.json, Custom Component Generator.json, Document Q&A.json, Meeting Summary.json, Memory Chatbot.json, Research Agent.json, Research Translation Loop.json, SEO Keyword Generator.json, Text Sentiment Analysis.json, Twitter Thread Generator.json, Vector Store RAG.json	Removed in-method provider-field visibility logic and imports (apply_provider_variable_config_to_build_config, get_language_model_options, update_model_options_in_build_config). Replaced update_build_config implementation to delegate to handle_model_input_update(self, build_config, field_value, field_name). Method signature unchanged; control flow simplified and centralized.
StructuredOutputComponent Update Refactoring src/backend/base/langflow/initial_setup/starter_projects/Financial Report Parser.json, Hybrid Search RAG.json, Image Sentiment Analysis.json	Updated imports to use handle_model_input_update instead of update_model_options_in_build_config and get_language_model_options. Replaced update_build_config implementation with single delegation call to handle_model_input_update, removing inline option caching and provider-specific field toggling.
AgentComponent Update Refactoring src/backend/base/langflow/initial_setup/starter_projects/Instagram Copywriter.json, Invoice Summarizer.json, Market Research.json, News Aggregator.json, Nvidia Remix.json, Pokédex Agent.json, Portfolio Website Code Generator.json, Price Deal Finder.json, SaaS Pricing.json, Search agent.json, Simple Agent.json, Social Media Agent.json	Replaced update_model_options_in_build_config with handle_model_input_update in update_build_config method. Several files also removed memory_inputs from AgentComponent public inputs. Updated code_hash metadata to reflect new implementation. Internal model option caching and field visibility now handled by centralized updater.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• fix: make ollama embeddings work with external URLs, make completion API come with type of last complete message #10558: Modifies LanguageModelComponent update_build_config across starter projects with async Ollama model fetching and related import changes, touching the same method-level code paths.
• fix: Handle params required for watsonX #10979: Changes unified_models get_llm signature to accept watsonx/ollama parameters and augments update_build_config with provider-specific visibility/requirement logic, affecting the same component update flow and import surface.

Suggested labels

refactor

Suggested reviewers

• edwinjosechittilappilly

---

Important

Pre-merge checks failed

Please resolve all errors before merging. Addressing warnings is optional.

❌ Failed checks (1 error, 1 warning)

Check name	Status	Explanation	Resolution
Test Coverage For New Implementations	❌ Error	PR modifies 20+ starter projects and multiple core components (base_component.py, csv_agent.py, openai_tools.py) with new centralized model handling functions, but no corresponding unit or integration test files were found.	Add test files for new functions (handle_model_input_update, get_embeddings) and integration tests verifying refactored components correctly delegate to centralized handlers.
Test Quality And Coverage	⚠️ Warning	No unit tests found for new handle_model_input_update and get_embeddings functions despite significant refactoring of model configuration logic across multiple components.	Add comprehensive pytest-based unit tests for both functions covering normal cases, error handling, provider-specific field visibility, option caching, and component integration.

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Test File Naming And Structure	✅ Passed	PR contains only JSON starter project file changes; no test files were modified or added.
Excessive Mock Usage Warning	✅ Passed	The PR demonstrates appropriate and well-designed mock usage rather than excessive mocking. Tests include real behavior verification and strategic mocking of external dependencies only.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The PR title clearly and specifically describes the main change: refactoring and unifying model input handling across components through centralized logic.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat-unify-models++

📝 Coding Plan

• Generate coding plan for human review comments

---

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.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

Frontend Unit Test Coverage Report

Coverage Summary

Lines	Statements	Branches	Functions
	24.38% (8698/35673)	17.21% (4831/28055)	17.06% (1275/7472)

Unit Test Results

Tests	Skipped	Failures	Errors	Time
2783	0 💤	0 ❌	0 🔥	45.596s ⏱️

[codecov]

Codecov Report

❌ Patch coverage is 44.44444% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 38.26%. Comparing base (7d4ffbc) to head (4071134).
⚠️ Report is 2 commits behind head on release-1.9.0.

Files with missing lines	Patch %	Lines
...base/langflow/agentic/services/flow_preparation.py	50.00%	4 Missing ⚠️
src/backend/base/langflow/api/utils/kb_helpers.py	0.00%	1 Missing ⚠️

❌ Your project status has failed because the head coverage (44.66%) is below the target coverage (60.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@                Coverage Diff                @@
##           release-1.9.0   #12025      +/-   ##
=================================================
+ Coverage          38.23%   38.26%   +0.02%     
=================================================
  Files               1647     1652       +5     
  Lines              82369    82457      +88     
  Branches           12223    12239      +16     
=================================================
+ Hits               31494    31551      +57     
- Misses             49092    49111      +19     
- Partials            1783     1795      +12     

Flag	Coverage Δ
backend	54.72% <44.44%> (-0.03%)	⬇️
frontend	21.76% <ø> (+<0.01%)	⬆️
lfx	44.66% <ø> (+0.07%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...nd/src/components/core/dropdownComponent/index.tsx	0.00% <ø> (ø)
...Component/components/modelInputComponent/index.tsx	67.88% <ø> (-0.24%)	⬇️
src/lfx/src/lfx/base/datastax/astradb_base.py	1.61% <ø> (ø)
.../lfx/base/models/google_generative_ai_constants.py	100.00% <ø> (ø)
...src/lfx/base/models/unified_models/build_config.py	27.27% <ø> (ø)
...c/lfx/base/models/unified_models/class_registry.py	31.25% <ø> (ø)
.../src/lfx/base/models/unified_models/credentials.py	9.25% <ø> (ø)
...rc/lfx/base/models/unified_models/instantiation.py	25.71% <ø> (ø)
...rc/lfx/base/models/unified_models/model_catalog.py	63.95% <ø> (ø)
...lfx/base/models/unified_models/provider_queries.py	68.00% <ø> (ø)
... and 3 more

... and 14 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.