agentjido / req_llm / e608a63c9f2e6ac486197ee29f5bef17d42255f5

49%

Build Type

push

github

Committed by web-flow

Commit Message

chore/refresh coverage tests (#70)

* feat(provider/google): add thinking budget option for Gemini 2.5 models

* Update Google and Groq provider tests to use Gemini 2.5 model and adjust max_tokens for Groq. Refresh captured_at timestamps in various fixture files for consistency. Add new fixture for Groq no tool call scenario.

* Enhance provider test helpers to validate response text content based on tool calls. Update various fixture files with new captured_at timestamps and add multiple new Google and XAI fixtures for tool call scenarios. Adjust existing fixture data for consistency across providers.

* Refactor Google provider usage calculation to handle missing candidatesTokenCount and update fixture data for consistency. Adjust captured_at timestamps and response IDs in multiple JSON fixture files.

* Moved integration tests into req_llm folder

* Normalize name of usage fixtures

* Refactor error handling in `req_llm.gen` mix task to improve clarity and functionality. Introduced dedicated functions for handling generation errors and missing API key errors, enhancing user feedback and guidance. Added checks for model not found errors and improved provider model listing logic.

* checkpoint

* Update .gitignore, refine valid providers list formatting, and adjust context limits in JSON fixtures for Anthropic and Vercel. Add cache_read cost to OpenCode JSON.

* Refactor Anthropic provider tests by consolidating coverage into a comprehensive test suite. Remove outdated core, object generation, streaming, tool calling, and usage tests. Update test models in configuration to reflect new naming conventions.

* Refactor streaming fixture handling by consolidating path management and updating fixture context setup. Remove outdated fixture functions and enhance test coverage for Google and Groq providers with new comprehensive test suites. Adjust existing fixture data for consistency and accuracy across multiple scenarios.

* Add Plug dependency for testin... (continued)

Run Details

910 of 2347 new or added lines in 42 files covered. (38.77%)

73 existing lines in 15 files now uncovered.

2663 of 5270 relevant lines covered (50.53%)

184.92 hits per line