q792602257 17a709c92f fix(perception): disable PaddleOCR doc-unwarping for screenshots ...

PaddleOCR's OCR.yaml pipeline defaults to use_doc_orientation_classify
and use_doc_unwarping enabled, which are meant for photographed paper
documents. Applied to a flat, upright device screenshot, UVDoc
geometrically warps the image before detection, and returns box
coordinates in that warped space with no inverse mapping back to the
original image.

Verified on a real screenshot: with unwarping on, the same detected
element ("新项目") shifts from y=158 to y=71 versus the original image,
and 2 boxes near the top edge (status bar time/battery) are dropped
entirely. Disabling both flags by default (still overridable via
explicit kwargs) makes detected boxes match the original screenshot.

2026-07-15 16:51:36 +08:00

..

__init__.py

Implement Apex Agent MVP scaffold

2026-07-06 13:22:15 +08:00

fakes.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00

test_agents_collab_runner.py

feat(runtime): capture step evidence in console

2026-07-15 10:12:09 +08:00

test_agents_e2e.py

multi agent

2026-07-06 23:23:44 +08:00

test_agents_integration.py

multi agent

2026-07-06 23:23:44 +08:00

test_agents_models.py

multi agent

2026-07-06 23:23:44 +08:00

test_agents_no_coupling.py

multi agent

2026-07-06 23:23:44 +08:00

test_agents_observer.py

multi agent

2026-07-06 23:23:44 +08:00

test_agents_reflector.py

multi agent

2026-07-06 23:23:44 +08:00

test_agents_verifier.py

multi agent

2026-07-06 23:23:44 +08:00

test_ai_planner_integration.py

feat(agent-runtime): add LLM-driven AI Planner with dual-provider tool calling

2026-07-12 13:48:50 +08:00

test_ai_planner_task_runner.py

feat: surface task execution progress across Host Agent and Cloud

2026-07-14 12:47:49 +08:00

test_ai_planner.py

feat(runtime): add planner reflection history with rationale and thinking

2026-07-15 12:43:22 +08:00

test_android_driver.py

feat(driver): add Android UiAutomator2 driver

2026-07-13 14:47:01 +08:00

test_android_integration.py

feat(driver): add Android UiAutomator2 driver

2026-07-13 14:47:01 +08:00

test_cloud_auth.py

feat(cloud): remove static credentials and add host console

2026-07-13 19:45:53 +08:00

test_cloud_client.py

feat(cloud-console): task listing, attempt history, CORS, and console SPA

2026-07-13 14:00:23 +08:00

test_cloud_composition_safety.py

feat(host-agent): make execution history authoritative

2026-07-15 11:46:27 +08:00

test_cloud_control_config.py

feat(cloud): remove static credentials and add host console

2026-07-13 19:45:53 +08:00

test_cloud_database.py

feat(cloud-store): support SQLite and PostgreSQL engines

2026-07-12 16:53:02 +08:00

test_cloud_dispatcher_real_task_runner.py

cloud

2026-07-06 23:44:18 +08:00

test_cloud_governance.py

feat(cloud): enforce host governance budgets

2026-07-13 22:56:31 +08:00

test_cloud_migrations.py

feat: surface task execution progress across Host Agent and Cloud

2026-07-14 12:47:49 +08:00

test_cloud_observability.py

feat(cloud-api): add correlated lifecycle logging

2026-07-12 18:37:51 +08:00

test_cloud_planner_decision_endpoint.py

feat: surface task execution progress across Host Agent and Cloud

2026-07-14 12:47:49 +08:00

test_cloud_repository_concurrency.py

test(cloud-scheduler): prove assignment concurrency

2026-07-12 17:42:20 +08:00

test_cloud_repository_contract.py

feat(runtime): add planner reflection history with rationale and thinking

2026-07-15 12:43:22 +08:00

test_cloud_sdk_api.py

feat: surface task execution progress across Host Agent and Cloud

2026-07-14 12:47:49 +08:00

test_cloud_skill_sync.py

feat(skills): cloud sync client + incremental sync + fork-on-revocation

2026-07-15 07:55:08 +08:00

test_cloud_store.py

fix(device-pool): scope primary key by host and thread through capability_tags

2026-07-07 08:30:47 +08:00

test_cloud_task_progress_visibility.py

feat: surface task execution progress across Host Agent and Cloud

2026-07-14 12:47:49 +08:00

test_cloud_user_auth.py

feat(cloud-console): add user authentication and administration

2026-07-13 17:54:53 +08:00

test_deployment_config.py

fix(cloud): align planner configuration and records

2026-07-15 09:43:14 +08:00

test_describe_screen_semantic.py

fix(semantic-scene): move describe_screen_semantic out of the tools/ layer

2026-07-07 08:30:49 +08:00

test_describe_screen.py

fix(perception): degrade to OCR-only when UI tree is unavailable

2026-07-15 10:03:30 +08:00

test_device_config.py

feat(cloud): add edge host enrollment

2026-07-13 13:54:16 +08:00

test_device_manager.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00

test_device_pool.py

fix(device-pool): scope primary key by host and thread through capability_tags

2026-07-07 08:30:47 +08:00

test_executor.py

Implement Apex Agent MVP scaffold

2026-07-06 13:22:15 +08:00

test_find_helpers.py

Implement Apex Agent MVP scaffold

2026-07-06 13:22:15 +08:00

test_host_agent_console_tasks.py

feat(host-agent): persist UI-tree evidence and add overlay/action visualization

2026-07-15 14:39:28 +08:00

test_host_agent_internal_api.py

docs(cloud): document governance rollout

2026-07-13 23:22:53 +08:00

test_host_agent_progress_reporting.py

feat: surface task execution progress across Host Agent and Cloud

2026-07-14 12:47:49 +08:00

test_host_agent_protocol_models.py

feat(host-protocol): define internal API models

2026-07-12 18:07:42 +08:00

test_host_agent_task_storage.py

feat(host-agent): make execution history authoritative

2026-07-15 11:46:27 +08:00

test_local_skills.py

feat(skills): open skill-management-console change + local skill store

2026-07-15 07:28:42 +08:00

test_mcp.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00

test_ocr.py

fix(perception): disable PaddleOCR doc-unwarping for screenshots

2026-07-15 16:51:36 +08:00

test_perception_provider.py

fix(perception): degrade to OCR-only when UI tree is unavailable

2026-07-15 10:03:30 +08:00

test_planner_config.py

feat(runtime): add planner reflection history with rationale and thinking

2026-07-15 12:43:22 +08:00

test_plugin_registry.py

fix(plugin-system): implement register_driver_type and stop discover() aborting on one bad manifest

2026-07-07 08:31:00 +08:00

test_scene_builder.py

fix(perception): reconcile points/pixels scale and stale overlay screenshot

2026-07-15 16:12:53 +08:00

test_semantic_enricher.py

fix(semantic-scene): preserve page/intents when all widgets are dangling, drop dead exception-name set

2026-07-07 08:31:03 +08:00

test_semantic_integration.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00

test_semantic_llm_client.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00

test_semantic_models.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00

test_semantic_task_loop.py

fix(semantic-scene): move describe_screen_semantic out of the tools/ layer

2026-07-07 08:30:49 +08:00

test_skill_catalog_e2e.py

feat(skills): unified read-merge + authoring/override MCP tools

2026-07-15 07:38:12 +08:00

test_skill_catalog_mcp.py

feat(skills): unified read-merge + authoring/override MCP tools

2026-07-15 07:38:12 +08:00

test_skill_catalog.py

feat(skill-catalog-subscription): synced catalog + HTTP sync + MCP tools

2026-07-07 10:33:23 +08:00

test_skill_retrieval.py

feat: add skill learning runtime

2026-07-06 18:07:53 +08:00

test_skill_store.py

fix(skill-versioning): consume divergence_tolerance, replace vacuous catalog-isolation guard

2026-07-07 08:31:11 +08:00

test_skill_sync_client.py

feat(skill-catalog-subscription): synced catalog + HTTP sync + MCP tools

2026-07-07 10:33:23 +08:00

test_skill_synthesis.py

feat: add skill learning runtime

2026-07-06 18:07:53 +08:00

test_skill_task_runner.py

feat: add skill learning runtime

2026-07-06 18:07:53 +08:00

test_skill_versioning.py

fix(skill-versioning): consume divergence_tolerance, replace vacuous catalog-isolation guard

2026-07-07 08:31:11 +08:00

test_smoke.py

cloud

2026-07-06 23:44:18 +08:00

test_task_dispatcher.py

cloud

2026-07-06 23:44:18 +08:00

test_task_loop.py

fix(perception): reconcile points/pixels scale and stale overlay screenshot

2026-07-15 16:12:53 +08:00

test_task_scheduler.py

feat(cloud): enforce host governance budgets

2026-07-13 22:56:31 +08:00

test_timeline.py

feat(host-agent): persist UI-tree evidence and add overlay/action visualization

2026-07-15 14:39:28 +08:00

test_tool_calling_client.py

fix(planner): allow rationale/thinking by using tool_choice=auto

2026-07-15 13:43:39 +08:00

test_tool_specs.py

feat(agent-runtime): add LLM-driven AI Planner with dual-provider tool calling

2026-07-12 13:48:50 +08:00

test_wda_integration.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00

test_workflow_conditions.py

workflow

2026-07-06 23:15:30 +08:00

test_workflow_models.py

workflow

2026-07-06 23:15:30 +08:00

test_workflow_runner.py

feat(host-agent): make execution history authoritative

2026-07-15 11:46:27 +08:00

test_workflow_skill_exec.py

workflow

2026-07-06 23:15:30 +08:00

test_workflow_store.py

workflow

2026-07-06 23:15:30 +08:00

test_workspace_packaging.py

Jenkins

2026-07-13 16:11:40 +08:00

test_world_model.py

fix(world-model): scope WorldModel state per-task to prevent cross-task contamination

2026-07-07 08:30:36 +08:00

test_world_models.py

feat(runtime): add planner reflection history with rationale and thinking

2026-07-15 12:43:22 +08:00

test_world_task_runner.py

feat: checkpoint device agent runtime milestones

2026-07-06 17:24:03 +08:00