chore(main): release 0.4.10

[github-actions]

🤖 I have created a release beep boop

0.4.10 (2026-09-17)

Features

• add ConfiguredProvider type (738f97c)
• add ConfiguredProvider type (c805397)
• added --watch to dev script (528ff4a)
• added a toggle for unloading models after test run (b7295c5)
• added auto downloading model from hf and deleting it after running the test (950fb6d)
• added average, min, max duration stats (35e516a)
• added average, min, max duration stats (5482c13)
• added BaseBadge component and disabled button state (35dfcf2)
• added BaseButton component (6cde7cd)
• added BaseRadioGroup component (d906862)
• added better support for authentication, configurable api key env variable names (6b0d031)
• added better support for authentication, configurable api key env variable names (b5acde8)
• added bleu evaluator (128cffa)
• added case details to test runs (3b9aa84)
• added centeralized context menu support (29b468f)
• added collapsing animation to external models (a04ef08)
• added confirm modal, deleting test runs and fixed stopping tests (ff12247)
• added context menus, navigation shortcuts and few refactors (8a14083)
• added cost and token usagedetails to test results (46ad46c)
• added custom error message for fetching models (811453f)
• added custom lm studio chat request (152a533)
• added custom scroll area component to fix scrollbar gutter issues (3840d29)
• added eval case sensivity (662b77b)
• added eval negation (c4e2024)
• added eval negation (e490fbb)
• added evaluation logic for exact match and contains (91a2b9e)
• added external provider support and implemented openrouter provider (a397e08)
• added file logging (7ad9a95)
• added hints and more information to several places (6c180cf)
• added input max length support (d2002ac)
• added json match evaluator (7a99b2f)
• added levenshtein and f1 evaluation methods (e1cf4b7)
• added lint to CI (cf76251)
• added local model action states (fd8391e)
• added logging level settings in ui (7bbe568)
• added logs view (cbaaaf3)
• added markdown rendered llm responses, collapsible options (11c7919)
• added min and max tok/s and TTFT (b368900)
• added model id and copy button to local model card (f04bf0b)
• added model load, unload and delete (48fe147)
• added more details to the current model run and a progress bar (d2c5912)
• added more options for chat requests (1d3c1ba)
• added multiple eval support for a test case (aa61401)
• added new shortcuts and shortcut composable (b43aac2)
• added not contains eval & fixed hf full url support (a654f5f)
• added ollama provider support (543f682)
• added onboarding view (738f97c)
• added onboarding view (f6ab522)
• added parallel run for external providers (8afb6c1)
• added playground (515c20a)
• added prompts view to manage system instructions seperately (fba4784)
• added provider auth support (0b0287a)
• added provider auth support (8fbb761)
• added provider sync ipc channel (ce19e01)
• added right click context menu to test case list (84ecccb)
• added rouge evaluation method (b03b91f)
• added score coloring and fixed some styles (26618df)
• added score coloring and fixed some styles (4f68baa)
• added semantic eval (3d4b82a)
• added setting to select LM Studio port and persisted settings in disk (e870daa)
• added sort order and better modality filters (74d051d)
• added split pane component (f26b3eb)
• added suite folder readonly setting (5d3271c)
• added support for custom eval pass threshold (643bbaf)
• added support for custom evaluator (5b63c46)
• added test case duration (4a0dc33)
• added thinking output to the results (a03168a)
• added toast component (e04b715)
• added tooltip support to field (cdb0da8)
• added tooltip support to modal (584fd6f)
• added vue devtools (c5d1e14)
• added vue devtools (1b0443c)
• added zoom keyboard shortcuts and mouse wheel (4efc5f9)
• auto save test suite (dc542f7)
• auto select loaded model and last ran suite when starting new test run (2fc6377)
• auto-fetch models when a provider is started via UI (39474f1)
• auto-select installed model if manual model input matches (d960ee0)
• better error handling and fixed passing chat options (8dbf3fa)
• complete ui redesign (ce9102d)
• complete ui redesign (c5da5a9)
• connected test runs page to list test suites (60e3b01)
• control lm studio server state in the settings view (2112674)
• default test suite config and context menu refactors (d12e516)
• display currently running test cases (f50f123)
• docs: added copy markdown, open in actions and updated docs (#66) (71e4653)
• edit provider configuration (738f97c)
• edit provider configuration (9ca8354)
• eval: added g-eval evaluation (91c6676)
• eval: added html validation evaluator (f8f15f4)
• eval: added llm-rubric evaluation (badf11d)
• expandable test input/output and copy button (e98c0f5)
• gh actions cache (849a991)
• global default test run settings (b8bff48)
• handle empty providers in test runs view (738f97c)
• handle empty providers in test runs view (324fd5c)
• implement base modal component and added new colors (302c051)
• implement initial test runner service (818ed93)
• implement LM Studio chat provider and IPC infrastructure (bd12c29)
• implement stopping tests and few button refactors (29ce019)
• implemented regex evaluator (3e7a4ef)
• improved test run cancelling and added timeout support (6dea077)
• initialize scss for base components (802dc7e)
• logs: reformatted log rows and updated some logs (e84fd74)
• models: added model downloads with quantization picker (#62) (9917e21)
• navigation: opened models view of the default provider (6b85e11)
• onboarding: used app icon and default provider view (4b577a0)
• output: added rendered HTML preview with custom template (6efc4c2)
• output: added rendered HTML preview with custom template (b6c2569)
• persist test runs and open save directory (ef8eacb)
• persistent window state (38dfcd2)
• prevent test runs from starting if we can't connect to the selected provider (07fff9d)
• provider: added api key probe to connection test (#64) (10640a9)
• provider: added extra request data passthrough (#54) (28e5b6b)
• provider: added managed server processes (37fab4e)
• provider: added openai provider with model metadata (#72) (b5e0735)
• provider: added support for llama.cpp (9676c45)
• provider: added support for llama.cpp (fb9bf8a)
• provider: added unsloth provider support (1aa7485)
• provider: added vercel ai gateway provider (c77da86)
• refactored settings view to be displayed as a modal (f6968de)
• registry (provider hub) model download support and some fixes (00f8e18)
• registry (provider hub) model download support and some fixes (8ab9686)
• replace hardcoded providers with user-managed configured provider collection (738f97c)
• replace hardcoded providers with user-managed configured provider collection (75171f8)
• replace native html titles with new tooltip component (84a52d5)
• save every setting to disk, better setting loading handler and fongiure server settings shortcut (7dc9ea6)
• server status loading effect (6e74abd)
• show current test and score while running a test suite (d4f7fcf)
• split logs to multiple files, refactor ipc channel registry and added more logs (942ea8b)
• storage: moved app data to .cannonade in home directory (#59) (8c93815)
• test-runs: added rerun button to run detail (7d64ea4)
• test-runs: toggle visibility of case detail fields (#55) (da0a805)
• test-suites: reclaimed editor space on suite detail page (#60) (5efe7e1)
• textarea resizer styling (62b9986)
• tooltip component and directive init (598186f)
• ui fixes, navigation refactor, context menu updates (b5a0c55)
• ui: judge settings and judge usage display (2ec1220)
• ui: raw json viewer for external models (0e1a6d7)
• ui: release date sort and display for external models (0de4179)
• unload models before running tests (1fc9ff4)
• updater: added auto updater with update badge and modal (#49) (c35d071)
• website: added a target and minor wording changes (e62f662)
• website: added hero background animation (7f36ee9)
• website: added macos setup modal and direct downloads (#70) (fa2a01e)
• website: added screenshots and fixed styles (17eb183)
• website: added screenshots and fixed styles (ab989f4)
• website: initial commit for website and docs (6fb27cf)
• website: website and docs (0cf80ab)

Bug Fixes

• added missing ollama chat options (7c701c4)
• added more vibrant icon (672686f)
• added more vibrant icon (59360a9)
• added write-file-atomic for safer file handling (481af59)
• allow multiple provider instances (738f97c)
• allow multiple provider instances (1a0bed1)
• ci: update workflow job names (6defbed)
• ci: update workflows (abc8d7d)
• ci: updated tsconfig (3c28ab4)
• ci: wrangler deployment fix (b0df5da)
• ci: wrangler deployment fix (ddd7a8d)
• code styling (b804da7)
• cosine similarity model loading on prod build (79dadec)
• cosine similarity model loading on prod build (ca91d03)
• docs: fixed base docs link (6740c6e)
• enabled sandbox and removed general purpose electronAPI (4bd1761)
• eval: aborted the model run when the judge throws an error (0f4c49e)
• eval: fixed case insensitive matching for turkish text (bb72ae2)
• eval: fixed llm-rubric threshold checks (a51a5d5)
• fix radiogroup borders and test run actions disappearing between states (c0dc47c)
• fixed added/removed test cases being shown in test runs (4654155)
• fixed aggregate stats idsplaying with zero values (8397842)
• fixed compiler deprecation warnings (61f4305)
• fixed keys for v-for (d446d82)
• fixed minor bugs and added tooltips (e3cb992)
• fixed screen flash during re-loading model listsand added updated time (9d81181)
• fixed time to first token mapping (7b1dc97)
• handled test suite and test run corrputed files (8f09fb0)
• Hide text overflow for long test suite names and descriptions (eae24c7)
• initial provider selection bug (00ab9f1)
• mac: fixed macOS installer (#53) (5a6e9ca)
• main: resolve executables from the login shell PATH (bdaa3b7)
• minor test fixes (fea9b33)
• minor ui fixes (a113e14)
• more space for test case editor (058aa56)
• new run model selector and run option bugs (e987f41)
• prevented reinitialization of test run listeners (8b0ee38)
• provider: improve llama.cpp download, load and router-mode errors (d3fa18e)
• remove grid from the actual output (019f9d3)
• removed vm2 dependency (b45b4a0)
• revert eval result border (639d995)
• revert ttft milliseconds change (4524d89)
• save test case footer align (4560d02)
• settings save debounce and initialization guard (abfdb7f)
• show only available capabilities for local models (39199d9)
• test-runs: fixed eval result cell widths for long details (1353cfc)
• tooltip: fixed tooltips never binding when content starts empty (a6ebfb2)
• ui fixes (6e8c199)
• ui: keep collapse height in sync with content during expand (9aadf83)
• ui: modal backdrop click fix (fe46535)
• ui: updated some styles and added model run transition (586d11b)
• ui: updated some styles and added model run transition (76ab5af)
• update tests (74f263c)
• updated tests (a5349a2)
• version bump, UI fixes and added removable prop to badge component (ab8a1f0)
• website: updated wording, styles and components (fab6bf6)

Performance

• website: improved animation performance (5904550)

Styling

• website: dropped mono font, reworked landing type, added seo meta tags (#63) (009e687)

Refactoring

• added field component (f183b9e)
• added input, number input and textarea components (03c006c)
• added panel component (5f75a55)
• added provider store (738f97c)
• added provider store (9a6f333)
• added toggle component (02f977c)
• base ui components and imports (6cfdcee)
• chevron and date format refactors (19fe9fc)
• converted styles to scss and style cleanup (5f51ce5)
• created chevron component (c329849)
• created model run row component (f9a4d7b)
• eval: renamed judge usage fields to judgeUsage (c56ccd2)
• flattened some meta fields and added provider capability check (51e0096)
• linter and test fixes (a2eecd0)
• migrated few modals to confirm modal component (dbcdb34)
• model store (451c607)
• moved animations to seperate file (70c3380)
• moved colors and variables to seperate scss files (abcccfc)
• moved copy button to seperate component (989df98)
• moved evals to electron side (1e90469)
• moved object mappers to separate files (26f6f74)
• provider interfaces and some impl (1b80b10)
• provider, preload and api files (869d935)
• refactored components (8804913)
• remove unused files (33bb725)
• renamed score variable (50daa16)
• renamed some components (8120d6b)
• seperate test runner to multiple files (68061e8)
• shared types (a36cd48)
• simplified toast (29069a4)
• test runner (7ea1acb)
• test-runs: extract evaluation summary formatting (4048e2a)
• test-runs: extract formatDurationMs and reformat metric displays (279e30d)
• ui and app ipc channel refactors (27bb73e)
• ui: extract Tooltip component and useTooltipPanel (4d4dcc5)
• update: added releases link, reused external link component (#65) (23267e0)
• updated custom components to use PascalCase (e57b689)
• updated folder structure for providers and shared files (7375fcb)
• updated the components (ffff3be)
• use badge component in model cards (50bc448)

Documentation

• readme: added screenshot and updated provider table (b280c08)
• skills: updated provider docs path and summarize rules (40fee51)
• website: added use cases to landing and docs (#71) (c8d1338)
• website: reworded use cases and added sampling settings (#74) (94a7ab5)
• website: rewrote guides and reference to match the app (9c28903)

Build System

• packaging: allowlisted packaged files per platform (f549498)

---

This PR was generated with Release Please. See documentation.