Release v1.12.6

[v1.12.6] - 2026-11-07

• test: simplify lifecycle contract and gateway operations

Release v1.12.5

[v1.12.5] - 2026-11-06

• Merge pull request #359 from tyemirov/bugfix/heygen-pilot-activation

Release v1.12.4

[v1.12.4] - 2026-10-25

• Merge pull request #355 from tyemirov/bugfix/document-explicit-manifest-inputs

Release v1.12.3

[v1.12.3] - 2026-10-23

• Merge pull request #352 from tyemirov/gix/add-xai-responses-reasoning-effort-none-for-grok-4-3

Release v1.12.2

[v1.12.2] - 2026-10-08

• fix: update release validator to mprlab.version-decision/v3

Release v1.12.0

[v1.12.0] - 2026-09-22

• Merge pull request #338 from tyemirov/gix/retain-browser-screenshots-only-for-failed-tests
• test: remove screenshot generation from UI specs
• test: enable screenshots on failure in Playwright config
• test: enable screenshot on failure in blackbox Playwright config
• chore: remove obsolete artifact screenshots
• docs: capture browser screenshots only on test failure
• docs: mark I285 complete - retain screenshots only on failure
• Merge pull request #337 from tyemirov/gix/keep-browser-screenshots-outside-tracked-artifacts
• test: attach screenshots via test.info outputPath instead of artifacts
• test: set Playwright outputDir to test-results/frontend
• test: set blackbox output directory and enable ts-check
• docs: clarify browser test screenshot output directories
• docs: mark B249 browser screenshot isolation as complete
• Merge pull request #336 from tyemirov/gix/filter-model-tasks-with-independently-selectable-buttons
• test: add e2e coverage for vision, audio and voice extraction tasks
• fix: require text input for vision, listen and voice extraction tasks
• docs: clarify required inputs for vision, listen and voice extraction
• chore: update I279 desktop screenshots
• docs: mark B247 and B248 as completed
• test: update specs for initially unselected task filters
• feat(ui): allow empty task filter and render task icons
• docs: clarify task filter and model icon behavior
• docs: update dashboard screenshot artifacts
• docs: mark B246, I284, F085 complete and add F086 spec
• fix: guard empty task selections and enforce AND offering matching
• test: run modelTasks unit tests as part of frontend:test
• docs: update dashboard artifact screenshots
• test(e2e): update management-ui assertions for task taxonomy
• feat: replace task flow strips with task icons on model cards
• refactor(public-site): use task icons in routing tree
• docs: clarify model card task icons in tenant connections
• chore(artifacts): update dashboard and explorer screenshots
• docs: mark I283 complete - show task icons on model cards
• test: update management-ui task assertions for compact labels
• style: simplify task picker and embed filter in models header
• docs(tenant-connections): clarify task filter button layout and behavior
• docs: refresh I279 and capability dashboard screenshots
• docs: mark I281 and I282 task selector improvements complete
• test: verify pressed task preserves selected model and unsaved prompt
• fix(connectionDashboard): skip render when task selection unchanged
• docs: mark B245 complete - keep model drafts on task reselect
• test: migrate dashboard and route explorer to task-filter selectors
• feat: replace capability domains with task-based model filtering
• feat(public): add task picker and modality filters to routing tree
• docs: document task filter toggles for tenant connections
• docs: add I279 dashboard screenshots and refresh mobile artifact
• docs: update README to describe task-based route explorer filtering
• docs: mark I279 and F084 complete and add I280 domain migration
• docs: add F081-F083 and I277-I279 to ISSUES.md
• Merge pull request #335 from tyemirov/gix/introduce-capability-domains-and-provider-routing
• ci: extend backend coverage job timeout to 15 minutes
• ci: split hosted backend into coverage and checks jobs
• test: stabilize high-load queue integration test
• build: forward ADMISSION_TEST_ARGS in test-upstream-admission-race
• docs: mark I275 queue saturation fix as complete
• test: cover ElevenLabs speech and voice conversion catalog
• feat: add speech_conversion capability and provider voice operations
• feat: add speech_conversion capability to public site renderer
• feat(python): expand supported asset MIME types for uploads
• feat: add speech conversion capability and audio media contract
• feat: add ElevenLabs speech generation and voice conversion support
• feat: add ElevenLabs speech generation and conversion
• feat: add ElevenLabs speech generation and speech conversion models
• docs: add provider speech contract reference to README
• chore: add test-provider-speech make target
• docs: mark B243, B244 and B242 complete and update F026 progress
• test(e2e): verify Vertex connection opens catalog API Keys page
• docs: add Vertex API key acquisition guide
• chore(providers): update Vertex AI key acquisition URL
• chore: document B242 Vertex API Keys page fix
• test: expand connection dashboard coverage for edge providers
• feat: extend provider catalog with services and resources
• feat(public-site): enhance renderer for provider services and controls
• feat: extend Python client with voice pagination and provider resources
• feat: add paginated voices, provider resources and aspect-ratio images
• feat(cli): add paginated voice filtering and strict model validation
• feat: add media-only verification and provider resource catalog
• feat: add FAL image operations and ElevenLabs provider resources
• feat: add FAL and ElevenLabs providers with Reve model and verification
• test: bump provider catalog schema version and expected provider count
• docs: clarify route explorer defaults and preflight capacity split
• chore: add make targets for provider resources and media tests
• docs: mark B237-B241, I273 and F079 complete, queue F080 and I274
• feat(proxy): migrate dictation defaults to transcription and speech
• docs: document B236 bounded capability defaults transfer
• docs: mark B236 transfer of retained capability defaults complete
• fix: make admission rejections definite and reject media-only dictation
• docs: clarify dictation operation for dictator transcription endpoints
• docs: document image admission rejection and transport uncertainty
• docs: mark B234 and B235 as complete
• feat(examples): add image generation client example
• chore(artifacts): add mobile capability dashboard image
• test: cover transcription split, upstream capacity and catalog growth
• feat: expand dashboard for image, transcription and speech domains
• feat: extend public catalog with image domains and upstream capacity
• feat(python): add media operation preview and lineage support
• feat: add image generation and editing media operations
• feat: add image generation pipeline and upstream admission control
• feat: add image generation pipeline and upstream admission control
• chore(deps): update Go module dependencies
• feat: add image generation APIs and upstream capacity admission
• feat: add upstream capacity and image generation support
• refactor(cli): migrate server config to upstream_capacity
• feat: add gpt-image generation/edit and upstream capacity admission
• build: add Makefile targets for upstream admission and image generation
• feat: add catalog-routed image generation and editing
• docs: mark B228-B233, I046, I269, I270, I272 as completed
• test: migrate management UI specs to auto-refresh and card selection
• feat: auto-refresh management dashboards and add capability domains
• test: update media command workflow to use whisper-base model
• refactor dictator media validation to be model-aware and domain-scoped
• docs: replace dictator-speech-v1 with whisper and TTS model enums
• feat: replace dictator-speech with whisper, qwen3 and silero models
• docs: update README for automatic usage refresh and I271 transition
• docs: archive 41 resolved issues after durable-documentation audit

Release v1.11.1

[v1.11.1] - 2026-09-17

• Merge pull request #334 from tyemirov/gix/wait-for-dashboard-startup-before-session-recovery
• docs: mark I268 resolved with dictator timeout isolation evidence
• test: isolate dictator acceptance operation timeouts
• docs: add I268 and I269 improvements to ISSUES.md
• test: wait for dashboard ready in shared UI migration spec
• test: wait for dashboard startup before session recovery [I267]
• Merge pull request #333 from tyemirov/gix/add-browser-cors-access-to-caddy-rate-limit-responses
• fix: add browser CORS to public-api rate limit responses

Release v1.11.0

[v1.11.0] - 2026-09-16

• Merge pull request #332 from tyemirov/gix/keep-dashboard-tenant-selected-at-all-times
• test: stabilize management-ui e2e profile request assertions with polling
• test: update management-ui e2e for always-selected dashboard tenant
• refactor(ui): remove Account usage and default to Default tenant
• docs: clarify dashboard tenant selection and usage scope
• docs: mark I265 superseded and I266 complete
• Merge pull request #331 from tyemirov/gix/expose-dictator-media-capabilities-through-catalog
• test: remove flaky timeout from dashboard connector scroll test
• Merge remote-tracking branch 'origin/master' into gix/expose-dictator-media-capabilities-through-catalog
• test: cover second speech provider and dashboard connector scrolling
• perf(connection-dashboard): throttle route drawing with rAF scheduling
• test(llmproxyclient): add mismatched closing delimiter case
• feat(llm-proxy-client): add media command for operations and assets
• feat(proxy): make dictator adapter provider-aware with extract duration
• feat: consolidate media gateway for dictator live and speech workflows
• docs: document dictator test targets and speech workflows
• build: add Makefile targets for media CLI and dictator live tests
• docs: sync issue tracker for B222-B225, I264 and F042
• Merge pull request #329 from tyemirov/improvement/device-validation-20260914
• docs: accept simulator validation and prohibit physical-device gates

Release v1.10.0

[v1.10.0] - 2026-09-15

• Merge pull request #330 from tyemirov/gix/harden-dictator-grpc-handling-and-add-provider
• test: add dictator protocol fixture for connection dashboard
• feat: add media capability support to connection dashboard
• feat(python): add provider diagnostics client and fix capabilities path
• feat: add provider diagnostics client and contract
• feat: add Dictator gRPC provider with account-bound media operations
• chore(deps): add Dictator speech SDK and gRPC dependencies
• chore(deps): add Dictator SDK and gRPC dependencies
• feat: add Dictator account connections and tenant provider diagnostics
• feat: convert dictator provider to tenant-owned connections
• docs(readme): clarify MediaOps retains only TelePrompter under F071
• chore: add make targets for dictator and provider diagnostics tests
• docs: update issue tracker for B218-B221, I263 and F065-F071
• Merge pull request #327 from tyemirov/tyemirov/improvement/I261-installed-gateway
• refactor: use installed Gateway runtime for lifecycle commands
• test: rebrand DashScope to Alibaba Cloud and update icon expectations
• chore(site): refresh branding icons and apple touch icon
• docs: rebrand DashScope Qwen to Alibaba Cloud in SEO resources
• refactor: enforce versionless account connection schema validation
• docs: rebrand DashScope to Alibaba Cloud and document schema transition
• chore(providers): rebrand DashScope as Alibaba Cloud
• refactor: adopt versionless schema and rename DashScope to Alibaba Cloud
• docs: update issue tracking and terminology for Alibaba Cloud
• test: verify provider prompt draft preservation on save failures
• fix(ui): preserve drafts for both connection dashboard forms on failure
• fix(client): use Bearer auth for asset uploads and enforce wait timeout
• fix(llmproxyclient): strip key query param from asset upload URL
• fix(proxy): retain active assets and recheck operation authority
• docs: clarify asset upload auth and media gateway operation guarantees
• test: expand Makefile test regexes to cover asset suites
• docs: mark B211-B217 bugfixes as complete
• chore(docker): update base image from bullseye to bookworm
• docs: update issue tracker for B209 and B210
• feat(media): add private Dictator adapter foundation (F042)
• feat(media): add durable media operations (F022)
• test(catalog): cover transport rejection paths (I258)
• refactor(catalog): compose transport components (I258)
• refactor(catalog): centralize protocol definitions (I257)
• test(protocols): add shared provider acceptance (I256)
• test: add connection dashboard coverage and migrate management specs
• refactor: replace provider management UI with connection dashboard
• docs: refresh SEO resources for connection dashboard
• feat: add account connections with idempotent creation support
• feat: introduce account-scoped provider connections
• feat: introduce account-owned provider connections
• chore: add Makefile targets for account connections and contract tests
• feat: implement tenant, connection and model dashboard

Release v1.9.1

[v1.9.1] - 2026-09-09

• Merge pull request #320 from tyemirov/tyemirov/I260-mpr-ui-migration
• I260: migrate shared UI config, footer menus, and session recovery