Robonix Webots CI Report

All 15 scenarios passed — VLM-primary scene graph + stable object identity

Provider: deepseek · Model: deepseek-chat · Generated On: 2026-07-05T12:55:20+00:00 · Stage: success · Confidence: high

All 15 CI scenarios passed (15/15, 100%). The PR introduces a VLM-primary scene graph where an image-grounded VLM call owns relational/semantic edges, while a fast geometric loop (3 Hz) handles only reachable_by. Object identity is stabilized via cross-tick re-bind, soft eviction with TTL, and new merge passes (same-class proximity, cross-class geometric collapse, configurable class groups). The pilot's RTDL parser now tolerates prose preamble around JSON. The scene-graph LLM enrichment ran but all VLM calls to the fake-vlm endpoint returned empty JSON, so no semantic edges were produced; the geometric loop and object registry functioned correctly.

PR Changes

system/pilot/src/planner.rs: Added extract_json_object() to strip prose preamble before JSON parsing; updated RTDL_PROTOCOL_REMINDER and retry prompt to require JSON-only replies; added unit tests for prose tolerance and JSON extraction.
system/scene/README.md: Documented VLM-primary scene graph, new env knobs (SCENE_CG_OBJ_MIN_POINTS, SCENE_CG_SAME_CLASS_MERGE_DIST_M, SCENE_CG_MERGE_CLASS_GROUPS, SCENE_OBJECT_TTL_SEC, SCENE_GRAPH_IMAGE_RELATIONS, VLM_REASONING_EFFORT), and troubleshooting for ghosting/label flicker.
system/scene/scene_service/ingest/perception_concept_graphs.py: Added same-class proximity merge, cross-class geometric collapse, configurable merge/identity gates via env, cross-tick re-bind in _apply_snapshot, soft eviction with TTL, and SCENE_CG_MERGE_CLASS_GROUPS parsing.
system/scene/scene_service/ingest/perception_vlm.py: Added VLM_REASONING_EFFORT env support forwarded to VLM request body.
system/scene/scene_service/scene_graph/builder.py: Added image-grounded relation dispatch (ImageRelationInferer), transient-blip guard preserving edges when <2 stable nodes, and split semantic/geometric edge storage.
system/scene/scene_service/scene_graph/geometric_loop.py: New file — fast 3 Hz loop producing reachable_by edges with debounce hysteresis.
system/scene/scene_service/scene_graph/geometry.py: New file — deterministic geometric predicates (on_top_of, under, inside, contains, near) and jitter-stable geometry signature for cache invalidation.
system/scene/scene_service/scene_graph/image_relations.py: New file — image-grounded VLM relation inference: projection, frame annotation, prompt, parsing.
system/scene/scene_service/scene_graph/llm_client.py: Added reasoning_effort support, multimodal image content in chat_json, and improved timeout logging.
system/scene/scene_service/scene_graph/prompts.py: Added SEMANTIC_RELATION_SYSTEM_PROMPT and build_semantic_relation_user_prompt for semantic-only inference.
system/scene/scene_service/scene_graph/relations.py: Added _normalize_relation, infer_semantic_relation, increased LLM timeout to 60s.
system/scene/scene_service/scene_graph/store.py: Split into geometric and semantic slices; added geometry-signature-based cache key, prune_relations, and compose get_snapshot.
system/scene/scene_service/scene_graph/types.py: Added reachable_by to RELATION_TYPES, geometric method to SceneGraphEdge.
system/scene/scene_service/service.py: Replaced RelationEngine with GeometricRelationLoop; wired scene-graph store and perception into builder.
system/scene/scene_service/state/__init__.py: Removed RelationEngine export.
system/scene/scene_service/state/object_registry.py: Added find_rebindable, soft_evict, prune_expired for cross-tick re-bind.
system/scene/scene_service/state/relations.py: Removed (replaced by geometric loop + image relations).
system/scene/scene_service/web.py: Removed on-canvas edge drawing; added explicit relation list in info panel.
system/scene/scripts/start.sh: Added env passthrough for new scene-graph knobs.
system/scene/tests/test_identity.py: New file — unit tests for find_rebindable, soft_evict, prune_expired, parse_merge_class_groups, cross-tick re-bind, TTL prune.
system/scene/tests/test_image_relations.py: New file — unit tests for projection, parsing, annotation, multimodal body, builder dispatch, transient-blip guard.
system/scene/tests/test_persistence.py: Added _object_ttl_s and _merge_class_group to detector stub.
system/scene/tests/test_scene_graph.py: Added tests for geometric_relation, compute_geometric_edges, geometry_signature, cache invalidation, store compose, debounce, infer_semantic_relation, normalize_relation, reasoning_effort, prune_relations.

Test Result

15 scenarios across builtin, cap, and flow families all passed. Builtin tests exercised file I/O, plan control, and command execution. Cap tests exercised camera/lidar snapshots, explore (timeout), map save, memory roundtrip, scene object listing, speech, and voiceprint. Flow tests exercised fault recovery (memory save), object navigation (explore → list objects → goal_near → navigate), and patrol observe (parallel camera/lidar → memory save → search). The scene-graph LLM enrichment ran but all VLM calls to the fake-vlm endpoint returned empty JSON, so no semantic edges were produced; the geometric loop and object registry functioned correctly.

Suggested Fix

No fix needed. The fake-vlm endpoint returns empty JSON, which is expected in CI without real VLM credentials. The scene-graph LLM enrichment correctly degrades to no semantic edges. If real VLM inference is desired, set VLM_BASE_URL/VLM_API_KEY/VLM_MODEL to a working endpoint.

Risks Or Watchouts

Scene-graph LLM enrichment produced no semantic edges because the fake-vlm endpoint returns empty JSON; this is expected in CI but means the VLM-primary path is untested end-to-end.
Object memory persistence failed (No module named 'pymilvus'), so warm-restore across restarts is disabled; this does not affect the current test run.
The explore skill's first attempt (5s timeout) waited for map→base_link tf and timed out without moving; the second attempt (8s timeout) successfully drove to a frontier and swept, then timed out. This is expected behavior for a bounded exploration test.
The scene-graph builder's transient-blip guard preserved edges when all objects went missing, but since no edges were ever produced, this path was not exercised.

Evidence

Source	Line	Excerpt
	5	"failed": 0,
	250	07-05 12:53:21.820 W scene [scene-graph-llm] request failed: JSONDecodeError: Expecting value: line 1 column 1 (char 0)
	4	failed — persistence OFF for this
	1	{"objects":[{"id":"scene.object.potted_plant_001","short_id":"potted_plant_001","cls":"potted_plant","pose":{"x":1.3788100595234163,"y":0.18861834707415223,"z":0.9535420492940094,"yaw":0.09825888521080683},"bbox":{"size_x":0.49310681558660924,"size_y":0.34721167011339704,"size_z":0.687749185640638,"yaw":0.09825888521080683},"confidence":0.688478285266507,"observation_count":34,"missing":false},{"id":"scene.object.robot_001","short_id":"robot_001","cls":"robot","pose":{"x":4.724330421092038e-10,"y":1.970265666254972e-15,"z":0.0,"yaw":5.259165653058311e-07},"bbox":{"size_x":0.6,"size_y":0.6,"size_z":1.5,"yaw":0.0},"confidence":1.0,"observation_count":21,"missing":false}],"relations":[],"scene_graph":{"edges":[],"updated_at":1783255896.606643},"robot":{"x":4.724330421092038e-10,"y":1.970265666254972e-15,"z":0.0,"yaw":5.259165653058311e-07},"occupancy":{"width":66,"height":51,"resolution":0.05000000074505806,"origin_x":-0.5249999761581421,"origin_y":-1.1249991655349731,"stamp_ms":178325589

Run Metadata

Test Environment

Scenarios

RTDL Plan Trees

Status	Suite	Scenario	Rounds	Dispatched contracts	Failures
PASS	builtin	fault_recovery_builtin	3	robonix/system/executor/builtin/read_file robonix/system/executor/builtin/write_file	-
PASS	builtin	file_roundtrip	1	robonix/system/executor/builtin/list_dir robonix/system/executor/builtin/read_file robonix/system/executor/builtin/write_file	-
PASS	builtin	plan_control_builtins	1	robonix/system/executor/builtin/get_all_plans robonix/system/executor/builtin/get_plan_status robonix/system/executor/builtin/stop_plan_at	-
PASS	builtin	run_command	1	robonix/system/executor/builtin/run_command	-
PASS	cap	camera_snapshot	1	robonix/primitive/camera/snapshot	-
PASS	cap	explore_smoke	2	robonix/skill/explore/explore robonix/skill/explore/explore/cancel robonix/skill/explore/explore/status	-
PASS	cap	lidar_snapshot	1	robonix/primitive/lidar/snapshot	-
PASS	cap	mapping_save	1	robonix/service/map/save_map	-
PASS	cap	memory_roundtrip	1	robonix/service/memory/save robonix/service/memory/search	-
PASS	cap	scene_object_fixture	1	robonix/system/scene/list_objects	-
PASS	cap	speech_speak	1	robonix/service/speech/speak	-
PASS	cap	voiceprint_list	1	robonix/service/voiceprint/list	-
PASS	flow	fault_recovery	4	robonix/service/memory/save	-
PASS	flow	object_navigation	4	robonix/service/navigation/navigate robonix/skill/explore/explore robonix/system/scene/goal_near robonix/system/scene/list_objects	-
PASS	flow	patrol_observe	3	robonix/primitive/camera/snapshot robonix/primitive/lidar/snapshot robonix/service/memory/save robonix/service/memory/search	-

Each scenario step is one VLM planning round. The tree below is the RTDL returned for that round; every do leaf has its own expected runtime contract and output checks.

PASSbuiltin/fault_recovery_builtin3 step(s), 3 leaf call(s)

step 0faulty read1 leaf

sequence faulty_read_step

do executor.builtin_read_file missing_read args {"path": "ci_never_created.txt"} expect robonix/system/executor/builtin/read_file failure

Observed plan round 0

FAIL1:0robonix/system/executor/builtin/read_fileargs {"path": "ci_never_created.txt"}No such file or directory (os error 2)

step 1recover by writing fallback1 leaf

sequence write_recovery_step

do executor.builtin_write_file write_recovery args {"content": "recovered charlie-3","path": "ci_recovered.txt"} expect robonix/system/executor/builtin/write_file success

Observed plan round 1

PASS2:0robonix/system/executor/builtin/write_fileargs {"content": "recovered charlie-3","path": "ci_recovered.txt"}wrote 19 bytes to /home/hanyulong/actions-runner/_work/robonix/robonix/examples/webots/ci_recovered.txt

step 2read fallback1 leaf

sequence read_recovery_step

do executor.builtin_read_file read_recovery args {"path": "ci_recovered.txt"} expect robonix/system/executor/builtin/read_file success

Observed plan round 2

PASS3:0robonix/system/executor/builtin/read_fileargs {"path": "ci_recovered.txt"}recovered charlie-3

PASSbuiltin/file_roundtrip1 step(s), 3 leaf call(s)

step 0write, read, and list probe file3 leafs

sequence file_roundtrip_tree

do executor.builtin_write_file write_probe args {"content": "robonix ci probe alpha-7","path": "ci_probe.txt"} expect robonix/system/executor/builtin/write_file success

do executor.builtin_read_file read_probe args {"path": "ci_probe.txt"} expect robonix/system/executor/builtin/read_file success

do executor.builtin_list_dir list_probe args {"path": "."} expect robonix/system/executor/builtin/list_dir success

Observed plan round 0

PASS1:0robonix/system/executor/builtin/write_fileargs {"content": "robonix ci probe alpha-7","path": "ci_probe.txt"}wrote 24 bytes to /home/hanyulong/actions-runner/_work/robonix/robonix/examples/webots/ci_probe.txt

PASS1:1robonix/system/executor/builtin/read_fileargs {"path": "ci_probe.txt"}robonix ci probe alpha-7

PASS1:2robonix/system/executor/builtin/list_dirargs {"path": "."}dir logs dir primitives dir rbnx-boot dir scripts dir services dir sim file .gitignore file README.md file ci_probe.txt file ci_recovered.txt file robonix_manifest.ci.generated.yaml file robonix_manifest.yaml file soma.yaml file start_rviz…

PASSbuiltin/plan_control_builtins1 step(s), 4 leaf call(s)

step 0inspect and stop current plan4 leafs

sequence inspect_and_stop_plan

do executor.builtin_get_all_plans list_active_plans args {} expect robonix/system/executor/builtin/get_all_plans success

do executor.builtin_get_plan_status inspect_current_plan args {"plan_id": "1"} expect robonix/system/executor/builtin/get_plan_status success

do executor.builtin_stop_plan_at stop_current_plan_on_complete args {"op_id": "1","plan_id": "1","when": "on_complete"} expect robonix/system/executor/builtin/stop_plan_at success

do executor.builtin_get_all_plans list_armed_stop_point args {} expect robonix/system/executor/builtin/get_all_plans success

Observed plan round 0

PASS1:0robonix/system/executor/builtin/get_all_plansargs {}{"count":1,"plans":[{"cancelled":false,"description":"inspect_and_stop_plan","op_count":5,"plan_id":"1","stop_points":0}]}

PASS1:1robonix/system/executor/builtin/get_plan_statusargs {"plan_id": "1"}{"cancelled":false,"ops":[{"description":"inspect_and_stop_plan","kind":"sequence","op_id":"16","state":"pending","stop_point":null},{"description":"list_active_plans","kind":"do","op_id":"17","state":"succeeded","stop_point":null},{"descr…

PASS1:2robonix/system/executor/builtin/stop_plan_atargs {"op_id": "1","plan_id": "1","when": "on_complete"}Stop point set: RTDL plan '1' will be cancelled on_complete op_id=1.

PASS1:3robonix/system/executor/builtin/get_all_plansargs {}{"count":1,"plans":[{"cancelled":false,"description":"inspect_and_stop_plan","op_count":5,"plan_id":"1","stop_points":1}]}

PASSbuiltin/run_command1 step(s), 1 leaf call(s)

step 0echo marker1 leaf

sequence echo_marker_step

do executor.builtin_run_command echo_marker args {"command": "echo robonix-ci-marker-bravo"} expect robonix/system/executor/builtin/run_command success

Observed plan round 0

PASS1:0robonix/system/executor/builtin/run_commandargs {"command": "echo robonix-ci-marker-bravo"}robonix-ci-marker-bravo

PASScap/camera_snapshot1 step(s), 1 leaf call(s)

step 0camera snapshot1 leaf

sequence camera_snapshot_step

do tiago_camera.camera_snapshot camera_snapshot args {} expect robonix/primitive/camera/snapshot success

Observed plan round 0

PASS1:0robonix/primitive/camera/snapshotargs {}{ "header": { "stamp": { "sec": 1783255897, "nanosec": 629281044 }, "frame_id": "head_front_camera_rgb_optical_frame" }, "height": 480, "width": 640, "encoding": "jpeg", "is_bigendian": 0, "step": 4451…

PASScap/explore_smoke2 step(s), 3 leaf call(s)

step 0start explore1 leaf

sequence start_explore_step

do explore.explore_explore start_explore args {"area_hint": "ci local room","max_speed_m_s": 0.1,"timeout_s": 5.0} expect robonix/skill/explore/explore failure

Observed plan round 0

FAIL1:0robonix/skill/explore/exploreargs {"area_hint": "ci local room","max_speed_m_s": 0.1,"timeout_s": 5.0}hit 5.0s ceiling

step 1poll and cancel explore2 leafs

sequence poll_cancel_explore

do explore.explore_status explore_status args {"run_id": ""} expect robonix/skill/explore/explore/status success

do explore.explore_cancel explore_cancel args {"run_id": ""} expect robonix/skill/explore/explore/cancel success

Observed plan round 1

PASS2:0robonix/skill/explore/explore/statusargs {"run_id": ""}{ "known": true, "state": "TIMEOUT", "area_m2": 0.9900000295042994, "frontiers_left": 1, "elapsed_s": 6.255239486694336, "eta_s": -1.0, "detail": "hit 5.0s ceiling" }

PASS2:1robonix/skill/explore/explore/cancelargs {"run_id": ""}{ "ok": true, "message": "task already in terminal state timeout (no-op)" }

PASScap/lidar_snapshot1 step(s), 1 leaf call(s)

step 0lidar snapshot1 leaf

sequence lidar_snapshot_step

do tiago_lidar.lidar_snapshot lidar_snapshot args {} expect robonix/primitive/lidar/snapshot success

Observed plan round 0

PASS1:0robonix/primitive/lidar/snapshotargs {}{ "header": { "stamp": { "sec": 99, "nanosec": 263716345 }, "frame_id": "hokuyo" }, "angle_min": -1.5700000524520874, "angle_max": 1.5700000524520874, "angle_increment": 0.0047647953033447266, "time_incr…

PASScap/mapping_save1 step(s), 1 leaf call(s)

step 0map save1 leaf

sequence map_save_step

do mapping.map_save_map map_save args {"map_id": "ci_test_map","note": "robonix ci"} expect robonix/service/map/save_map success

Observed plan round 0

PASS1:0robonix/service/map/save_mapargs {"map_id": "ci_test_map","note": "robonix ci"}{ "ok": true, "map_id": "ci_test_map", "database_path": "/mapping/maps/ci_test_map/rtabmap.db", "detail": "saved ci_test_map → /mapping/maps/ci_test_map" }

PASScap/memory_roundtrip1 step(s), 2 leaf call(s)

step 0memory save + search2 leafs

sequence memory_save_search

do memory.memory_save save_fact args {"data": "ci-fact: the test beacon is alpha-7"} expect robonix/service/memory/save success

do memory.memory_search search_fact args {"data": "test beacon"} expect robonix/service/memory/search success

Observed plan round 0

PASS1:0robonix/service/memory/saveargs {"data": "ci-fact: the test beacon is alpha-7"}{ "data": "Memory saved and indexed." }

PASS1:1robonix/service/memory/searchargs {"data": "test beacon"}{ "data": "Relevant memories:\n- ci-fact: the test beacon is alpha-7" }

PASScap/scene_object_fixture1 step(s), 1 leaf call(s)

step 0scene list objects1 leaf

sequence scene_objects_step

do scene.scene_list_objects list_scene_objects args {} expect robonix/system/scene/list_objects success

Observed plan round 0

PASS1:0robonix/system/scene/list_objectsargs {}{ "objects": [ { "id": "scene.object.potted_plant_001", "label": "potted_plant", "x": 1.401408995854423, "y": 0.18870096755765672, "z": 0.9376586536325674, "yaw": 0.11042965140630168, "last_s…

PASScap/speech_speak1 step(s), 1 leaf call(s)

step 0speak1 leaf

sequence speech_speak_step

do speech.speech_speak speech_speak args {"target": "","text": "robonix continuous integration test"} expect robonix/service/speech/speak success

Observed plan round 0

PASS1:0robonix/service/speech/speakargs {"target": "","text": "robonix continuous integration test"}{ "ok": true, "detail": "spoke 35 chars on audio_driver" }

PASScap/voiceprint_list1 step(s), 1 leaf call(s)

step 0voiceprint list1 leaf

sequence voiceprint_list_step

do voiceprint.voiceprint_list voiceprint_list args {} expect robonix/service/voiceprint/list success

Observed plan round 0

PASS1:0robonix/service/voiceprint/listargs {}{ "users_json": "[]", "count": 0, "error": "db not initialised" }

PASSflow/fault_recovery2 step(s), 2 leaf call(s)

step 0faulty save1 leaf

sequence faulty_save_step

do memory.memory_save faulty_save args {} expect robonix/service/memory/save failure

Observed plan round 0

FAIL1:0robonix/service/memory/saveargs {}mcp tool error: Error executing tool save: 1 validation error for saveArguments data Field required [type=missing, input_value={}, input_type=dict] For further information visit https://errors.pydantic.dev/2.13/v/missing

step 1recover1 leaf

sequence recovery_save_step

do memory.memory_save recovery_save args {"data": "ci-flow recovered fact: beacon bravo-2"} expect robonix/service/memory/save success

Observed plan round 1

PASS2:0robonix/service/memory/saveargs {"data": "ci-flow recovered fact: beacon bravo-2"}{ "data": "Memory saved and indexed." }

PASSflow/object_navigation4 step(s), 4 leaf call(s)

step 0explore before object navigation1 leaf

sequence explore_before_navigation

do explore.explore_explore start_explore args {"area_hint": "ci local room","max_speed_m_s": 0.1,"timeout_s": 8.0} expect robonix/skill/explore/explore failure

Observed plan round 0

FAIL1:0robonix/skill/explore/exploreargs {"area_hint": "ci local room","max_speed_m_s": 0.1,"timeout_s": 8.0}hit 8.0s ceiling

step 1list scene objects during explored state1 leaf

sequence select_scene_object

do scene.scene_list_objects list_scene_objects args {} expect robonix/system/scene/list_objects success

Observed plan round 1

PASS2:0robonix/system/scene/list_objectsargs {}{ "objects": [ { "id": "scene.object.potted_plant_001", "label": "potted_plant", "x": 1.4364588563622949, "y": 0.2083433256283281, "z": 0.9956578563902584, "yaw": -0.23948391417064885, "last_…

step 2resolve detected object approach pose1 leaf

sequence resolve_approach_pose

do scene.scene_goal_near goal_near_object args {"object_id": {"var": "object_id"}} expect robonix/system/scene/goal_near success

Observed plan round 2

PASS3:0robonix/system/scene/goal_nearargs {"object_id": "scene.object.potted_plant_001"}{ "reachable": true, "x": 0.5372809373518329, "y": 0.1698846469716823, "yaw": 0.04274488069038888, "reason": "approach pose for 'potted_plant' (scene.object.potted_plant_001)" }

step 3navigate to object approach pose1 leaf

sequence navigate_to_object

do nav2.navigation_navigate navigate_object_goal args {"goal": {"header": {"frame_id": "map","stamp": {"nanosec": 0,"sec": 0}},"pose": {"orientation": {"w": {"transform": "cos_half","var": "nav_yaw"},"x": 0.0,"y": 0.0,"z": {"transform": "sin_half","var": "nav_yaw"}},"position": {"x": {"var": "nav_x"},"y": {"var": "nav_y"},"z": 0.0}}}} expect robonix/service/navigation/navigate success

Observed plan round 3

PASS4:0robonix/service/navigation/navigateargs {"goal": {"header": {"frame_id": "map","stamp": {"nanosec": 0,"sec": 0}},"pose": {"orientation": {"w": 0.9997716180904244,"x": 0.0,"y": 0.0,"z": 0.02137081329417476},"position": {"x": 0.5372809373518329,"y": 0.1698846469716823,"z": 0.0}}}}{ "known": true, "state": "SUCCEEDED", "detail": "succeeded" }

PASSflow/patrol_observe3 step(s), 4 leaf call(s)

step 0parallel sensor capture2 leafs

parallel observe_step

do tiago_camera.camera_snapshot observe_snapshot args {} expect robonix/primitive/camera/snapshot success

do tiago_lidar.lidar_snapshot observe_lidar args {} expect robonix/primitive/lidar/snapshot success

Observed plan round 0

PASS1:0robonix/primitive/camera/snapshotargs {}{ "header": { "stamp": { "sec": 1783256054, "nanosec": 481575250 }, "frame_id": "head_front_camera_rgb_optical_frame" }, "height": 480, "width": 640, "encoding": "jpeg", "is_bigendian": 0, "step": 4451…

PASS1:1robonix/primitive/lidar/snapshotargs {}{ "header": { "stamp": { "sec": 211, "nanosec": 544298000 }, "frame_id": "hokuyo" }, "angle_min": -1.5700000524520874, "angle_max": 1.5700000524520874, "angle_increment": 0.0047647953033447266, "time_inc…

step 1remember1 leaf

sequence remember_step

do memory.memory_save remember_observation args {"data": "ci-flow observation: snapshot captured at waypoint 1"} expect robonix/service/memory/save success

Observed plan round 1

PASS2:0robonix/service/memory/saveargs {"data": "ci-flow observation: snapshot captured at waypoint 1"}{ "data": "Memory saved and indexed." }

step 2recall1 leaf

sequence recall_step

do memory.memory_search recall_observation args {"data": "waypoint 1"} expect robonix/service/memory/search success

Observed plan round 2

PASS3:0robonix/service/memory/searchargs {"data": "waypoint 1"}{ "data": "Relevant memories:\n- ci-fact: the test beacon is alpha-7\n\nci-flow recovered fact: beacon bravo-2\n\nci-flow recovered fact: beacon bravo-2\n\nci-flow recovered fact: beacon bravo-2\n\nci-flow observation: snapshot captured …

Logs

Build stdout is embedded as sim_logs/build-webots-deployment.log when the build step starts. The deploy supervisor output with the Robonix boot banner is embedded as rbnx-boot.log.

Runtime Contract Coverage

Contracts observed by the scenario runner in executor leaf results. This shows which runtime contract ids were exercised by this run; it is not a full API inventory.

Generated On	2026-07-05T20:55:20+08:00
Test Report Generated On	2026-07-05T20:54:20+08:00
Run Id	28741232169
Workflow	Robonix CI Bot
Event	issue_comment
Repository	syswonder/robonix
Tested Commit	f5ddd1c08c14
Tested Commit Subject	Merge remote-tracking branch 'refs/remotes/robonix-pr/116/head' into dev
Tested Commit Author	github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Tested Commit Date	2026-07-05T20:47:53+08:00
Pr Number	#116
Pr Title	feat(scene): VLM-primary scene graph + stable object identity
Pr Author	HeartLinked
Pr Url	https://github.com/syswonder/robonix/pull/116
Pr Head Ref	dev-sc-rebase
Pr Head Repo	HeartLinked/robonix
Triggered By	enkerewpo
Actor	enkerewpo
Triggering Actor	enkerewpo
Runner Name	robonix-runner-pku-rtx5090-x64
Runner Os	Linux
Runner Arch	X64
Origin Url	https://github.com/syswonder/robonix
Run Attempt	1
Ref	refs/heads/dev
Sha	03528d89c627
Pr Base Ref	dev
Pr Head Sha	0887c97ecf091807a5a1a49d16b46dc34ac5215a
Report Id	28741232169
Test Ref	refs/heads/dev

Hostname	user-NF5468M6
Os Pretty Name	Ubuntu 22.04.5 LTS
Kernel	Linux 6.8.0-124-generic x86_64 GNU/Linux
Arch	x86_64
Cpu Model	Intel(R) Xeon(R) Platinum 8368Q CPU @ 2.60GHz
Cpu Cores	152
Mem Total	125Gi
Gpu Summary	0, NVIDIA GeForce RTX 5090, 32607 MiB, 580.65.06;1, NVIDIA GeForce RTX 5090, 32607 MiB, 580.65.06;2, NVIDIA GeForce RTX 5090, 32607 MiB, 580.65.06;3, NVIDIA GeForce RTX 5090, 32607 MiB, 580.65.06
Nvidia Driver	580.65.06
Cuda Version	13.0
Docker Version	29.2.1
Docker Compose Version	5.0.2
Webots Version	webots package 2025a
Ros Distro	humble
Ros Domain Id	159
Sim Container	ci-28741232169-sim
Sim Image	ci-28741232169-sim
Webots Stream Port	23169
Webots Viewer Port	24169
Python Version	Python 3.10.12
Rustc Version	rustc 1.95.0 (59807616e 2026-04-14)
Rbnx Version	rbnx 0.1.0

Area	Contract id
robonix/primitive/camera	robonix/primitive/camera/snapshot
robonix/primitive/lidar	robonix/primitive/lidar/snapshot
robonix/service/map	robonix/service/map/save_map
robonix/service/memory	robonix/service/memory/save
robonix/service/memory	robonix/service/memory/search
robonix/service/navigation	robonix/service/navigation/navigate
robonix/service/speech	robonix/service/speech/speak
robonix/service/voiceprint	robonix/service/voiceprint/list
robonix/skill/explore	robonix/skill/explore/explore
robonix/skill/explore	robonix/skill/explore/explore/cancel
robonix/skill/explore	robonix/skill/explore/explore/status
robonix/system/executor	robonix/system/executor/builtin/get_all_plans
robonix/system/executor	robonix/system/executor/builtin/get_plan_status
robonix/system/executor	robonix/system/executor/builtin/list_dir
robonix/system/executor	robonix/system/executor/builtin/read_file
robonix/system/executor	robonix/system/executor/builtin/run_command
robonix/system/executor	robonix/system/executor/builtin/stop_plan_at
robonix/system/executor	robonix/system/executor/builtin/write_file
robonix/system/scene	robonix/system/scene/goal_near
robonix/system/scene	robonix/system/scene/list_objects

Robonix Webots CI Report

LLM-Assisted Analysis

PR Changes

Test Result

Suggested Fix

Risks Or Watchouts

Evidence

Run Metadata

Test Environment

Scenarios

RTDL Plan Trees

Logs

Runtime Contract Coverage