Add transport config to blueprints

[Dreamsorcerer]

This merges into #2048. Moves transport config to be part of the general blueprint config.

[codecov]

❌ 2 Tests Failed:

Tests completed	Failed	Passed	Skipped
2158	2	2156	71

View the full list of 2 ❄️ flaky test(s)

dimos.e2e_tests.test_dimsim_path_replaning::test_path_replanning

Flake rate in main: 7.14% (Passed 13 times, Failed 1 times)

Stack Traces | 242s run time

lcm_spy = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d521f020>
start_blueprint = <function start_blueprint.<locals>.set_name_and_start at 0x7198d54c4360>
dim_sim = <dimos.e2e_tests.dim_sim_client.DimSimClient object at 0x7198d55d8d10>
direct_cmd_vel_explorer = <dimos.simulation.mujoco.direct_cmd_vel_explorer.DirectCmdVelExplorer object at 0x7198d5533d10>
spawn_wall_on_pose = <function spawn_wall_on_pose.<locals>.spawn at 0x7198d54c49a0>

    @pytest.mark.self_hosted_large
    def test_path_replanning(
        lcm_spy, start_blueprint, dim_sim, direct_cmd_vel_explorer, spawn_wall_on_pose
    ) -> None:
        start_blueprint(
            "--dimsim-scene=empty",
            "run",
            "unitree-go2-agentic",
            simulator="dimsim",
        )
        lcm_spy.save_topic(".../McpClient/on_system_modules/res")
        lcm_spy.wait_for_saved_topic(".../McpClient/on_system_modules/res", timeout=1200.0)
    
        # robot spawns at (3, 2)
    
        # side wall
        dim_sim.add_wall(2, -2.5, 12, -2.5)
        # other side wall
        dim_sim.add_wall(2, 3.5, 12, 3.5)
        # back wall (behind robot)
        dim_sim.add_wall(2, -2.5, 2, 3.5)
        # forward wall (far end)
        dim_sim.add_wall(12, -2.5, 12, 3.5)
        # dividing wall at x=7 with doors at y=[-1.5,-0.5] and y=[1.5,2.5]
        dim_sim.add_wall(7, -2.5, 7, -1.5)
        dim_sim.add_wall(7, -0.5, 7, 1.5)
        dim_sim.add_wall(7, 2.5, 7, 3.5)
    
        direct_cmd_vel_explorer.linear_speed = 0.8
        direct_cmd_vel_explorer.follow_points([(10, 2), (2.5, 2), (3, 2)])
    
        # When the robot comes within 1.5 m of the left door's centre, drop a wall
        # in the opening so the planner has to bail out and route through the
        # right door at y=-1 instead.
        spawn_wall_on_pose(
            point=(7, 2),
            threshold=1.5,
            wall=(7, 1.5, 7, 2.5),
        )
    
        dim_sim.publish_goal(10.913, 0.588)
    
>       lcm_spy.wait_until_odom_position(10.913, 0.588, threshold=1, timeout=120)

dim_sim    = <dimos.e2e_tests.dim_sim_client.DimSimClient object at 0x7198d55d8d10>
direct_cmd_vel_explorer = <dimos.simulation.mujoco.direct_cmd_vel_explorer.DirectCmdVelExplorer object at 0x7198d5533d10>
lcm_spy    = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d521f020>
spawn_wall_on_pose = <function spawn_wall_on_pose.<locals>.spawn at 0x7198d54c49a0>
start_blueprint = <function start_blueprint.<locals>.set_name_and_start at 0x7198d54c4360>

dimos/e2e_tests/test_dimsim_path_replaning.py:60: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
dimos/e2e_tests/lcm_spy.py:182: in wait_until_odom_position
    self.wait_for_message_result(
        predicate  = <function LcmSpy.wait_until_odom_position.<locals>.predicate at 0x7198d54c5300>
        self       = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d521f020>
        threshold  = 1
        timeout    = 120
        x          = 10.913
        y          = 0.588
dimos/e2e_tests/lcm_spy.py:168: in wait_for_message_result
    self.wait_until(
        event      = <threading.Event at 0x7198d4d31df0: unset>
        fail_message = 'Failed to get to position x=10.913, y=0.588'
        listener   = <function LcmSpy.wait_for_message_result.<locals>.listener at 0x7198d54c53a0>
        predicate  = <function LcmSpy.wait_until_odom_position.<locals>.predicate at 0x7198d54c5300>
        self       = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d521f020>
        timeout    = 120
        topic      = '/odom#geometry_msgs.PoseStamped'
        type       = <class 'dimos.msgs.geometry_msgs.PoseStamped.PoseStamped'>
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d521f020>

    def wait_until(
        self,
        *,
        condition: Callable[[], bool],
        timeout: float,
        error_message: str,
        poll_interval: float = 0.1,
    ) -> None:
        start_time = time.time()
        while time.time() - start_time < timeout:
            if condition():
                return
            time.sleep(poll_interval)
>       raise TimeoutError(error_message)
E       TimeoutError: Failed to get to position x=10.913, y=0.588

condition  = <bound method Event.is_set of <threading.Event at 0x7198d4d31df0: unset>>
error_message = 'Failed to get to position x=10.913, y=0.588'
poll_interval = 0.1
self       = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d521f020>
start_time = 1781808335.9697042
timeout    = 120

dimos/e2e_tests/lcm_spy.py:105: TimeoutError

dimos.e2e_tests.test_dimsim_walk_forward::test_walk_forward

Flake rate in main: 33.33% (Passed 12 times, Failed 6 times)

Stack Traces | 206s run time

lcm_spy = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d4d33020>
start_blueprint = <function start_blueprint.<locals>.set_name_and_start at 0x7198d54c59e0>
human_input = <function human_input.<locals>.send_human_input at 0x7198d54c5e40>
dim_sim = <dimos.e2e_tests.dim_sim_client.DimSimClient object at 0x7198d4d54d40>

    @pytest.mark.self_hosted_large
    def test_walk_forward(lcm_spy, start_blueprint, human_input, dim_sim) -> None:
        start_blueprint(
            "run",
            "--disable",
            "spatial-memory",
            "--disable",
            "security-module",
            "unitree-go2-agentic",
            simulator="dimsim",
        )
        lcm_spy.save_topic(".../McpClient/on_system_modules/res")
        lcm_spy.wait_for_saved_topic(".../McpClient/on_system_modules/res", timeout=1200.0)
    
        origin_x, origin_y = 1, 2
        dim_sim.set_agent_position(origin_x, origin_y)
    
        human_input("move forward 3 meter")
    
>       lcm_spy.wait_until_odom_position(origin_x + 3, origin_y, threshold=0.4, timeout=120)

dim_sim    = <dimos.e2e_tests.dim_sim_client.DimSimClient object at 0x7198d4d54d40>
human_input = <function human_input.<locals>.send_human_input at 0x7198d54c5e40>
lcm_spy    = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d4d33020>
origin_x   = 1
origin_y   = 2
start_blueprint = <function start_blueprint.<locals>.set_name_and_start at 0x7198d54c59e0>

dimos/e2e_tests/test_dimsim_walk_forward.py:37: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
dimos/e2e_tests/lcm_spy.py:182: in wait_until_odom_position
    self.wait_for_message_result(
        predicate  = <function LcmSpy.wait_until_odom_position.<locals>.predicate at 0x7198d54c5a80>
        self       = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d4d33020>
        threshold  = 0.4
        timeout    = 120
        x          = 4
        y          = 2
dimos/e2e_tests/lcm_spy.py:168: in wait_for_message_result
    self.wait_until(
        event      = <threading.Event at 0x7198d4d542f0: unset>
        fail_message = 'Failed to get to position x=4, y=2'
        listener   = <function LcmSpy.wait_for_message_result.<locals>.listener at 0x7198d54c5bc0>
        predicate  = <function LcmSpy.wait_until_odom_position.<locals>.predicate at 0x7198d54c5a80>
        self       = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d4d33020>
        timeout    = 120
        topic      = '/odom#geometry_msgs.PoseStamped'
        type       = <class 'dimos.msgs.geometry_msgs.PoseStamped.PoseStamped'>
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d4d33020>

    def wait_until(
        self,
        *,
        condition: Callable[[], bool],
        timeout: float,
        error_message: str,
        poll_interval: float = 0.1,
    ) -> None:
        start_time = time.time()
        while time.time() - start_time < timeout:
            if condition():
                return
            time.sleep(poll_interval)
>       raise TimeoutError(error_message)
E       TimeoutError: Failed to get to position x=4, y=2

condition  = <bound method Event.is_set of <threading.Event at 0x7198d4d542f0: unset>>
error_message = 'Failed to get to position x=4, y=2'
poll_interval = 0.1
self       = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7198d4d33020>
start_time = 1781809026.7287138
timeout    = 120

dimos/e2e_tests/lcm_spy.py:105: TimeoutError

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[greptile-apps]

Greptile Summary

This PR moves WebRTC transport configuration into the standard blueprint config schema, so credentials and transport settings can be supplied via -o transports.<name>.<field>=... CLI flags or TRANSPORTS__<NAME>__<FIELD>=... env vars instead of the old transport-specific env vars (TELEOP_*, CF_TELEOP_*). As part of this, ProviderConfig and its subclasses are migrated from @dataclass(frozen=True) to Pydantic BaseModel with frozen=True, extra="forbid".

• Blueprint.config() dynamically appends a transports sub-model to the schema by introspecting each transport's _config_cls; _apply_transport_overrides in the coordinator then rebuilds affected transports with the merged config before startup.
• BrokerConfig and CloudflareConfig have their TELEOP_* / CF_TELEOP_* env-var fallbacks unconditionally removed, making this a hard breaking change for any deployment that relied on those names.
• with_config_overrides on both WebRTCTransport and WebRTCVideoTransport uses model_copy(update=...), which in Pydantic v2 does not coerce types, so non-string fields (ordered: bool, heartbeat_hz: float, max_retransmits: int | None) will hold raw strings when the override originates from the CLI or an env var.

Confidence Score: 3/5

The core config-routing wiring is structurally sound, but two transports silently accept raw-string values for typed fields, and the hard removal of legacy env vars will break existing deployments without warning.

Both with_config_overrides implementations use model_copy(update=...) without re-validation. CLI flags and env vars arrive as strings, so fields like ordered: bool and heartbeat_hz: float end up holding raw strings that raise TypeError when used. Additionally, every deployment using the old TELEOP_API_KEY / CF_TELEOP_APP_ID env var family will hit a RuntimeError at startup with no deprecation grace period.

dimos/core/transport.py — both with_config_overrides implementations need model_validate instead of model_copy; dimos/protocol/pubsub/impl/webrtc/providers/broker.py and cloudflare.py for the breaking env-var removal.

Important Files Changed

Filename	Overview
dimos/core/coordination/blueprints.py	Adds dynamic transports.* sub-model to the blueprint config schema. Logic is clean: deduplication via seen set and the transport_config_name helper are straightforward. No issues found.
dimos/core/coordination/module_coordinator.py	Extracts transports overrides from blueprint_args and applies them via _apply_transport_overrides. Raw string values from CLI/env are forwarded to with_config_overrides without re-validation, causing type mismatches for non-string fields.
dimos/core/transport.py	Implements with_config_overrides on WebRTCTransport and WebRTCVideoTransport using model_copy(update=...), which does not coerce types in Pydantic v2, leaving non-string fields as raw strings when overrides originate from CLI or env.
dimos/protocol/pubsub/impl/webrtc/providers/spec.py	Migrates ProviderConfig from @DataClass(frozen=True) to Pydantic BaseModel. Equality and hashing semantics preserved; singleton _providers dict continues to work correctly.
dimos/protocol/pubsub/impl/webrtc/providers/broker.py	Migrates BrokerConfig to Pydantic, hardens defaults, and removes all TELEOP_* env-var fallbacks — a hard breaking change for existing deployments.
dimos/protocol/pubsub/impl/webrtc/providers/cloudflare.py	Migrates CloudflareConfig to Pydantic, removes CF_TELEOP_* env-var fallbacks, and exports MAX_MSG_SIZE for the benchmark. Same breaking-change concern as broker.py.
dimos/robot/cli/dimos.py	Fixes arg_help to handle transport config nodes that have no backing BlueprintAtom. Null-safe lookup and guarded _atom access are correct.
dimos/protocol/pubsub/impl/webrtc/test_transport.py	Adds thorough new tests for Pydantic frozen semantics, blueprint config schema exposure, and transport override rebuild. Removes now-unnecessary monkeypatching of deleted env vars.
dimos/teleop/quest_hosted/blueprints.py	Updates run comments to reflect the new CLI syntax. No code changes.

Comments Outside Diff (1)

1. dimos/protocol/pubsub/impl/webrtc/providers/broker.py, line 64-74 (link)

   Breaking removal of TELEOP_* env vars

   The previous implementation fell back to TELEOP_API_KEY, TELEOP_BROKER_URL, TELEOP_ROBOT_ID, and TELEOP_ROBOT_NAME. All four are now silently dropped. Any deployment that sets those env vars will fail at BrokerProvider construction with a runtime error (not a helpful deprecation warning). The same applies to CF_TELEOP_APP_ID / CF_TELEOP_APP_SECRET in CloudflareConfig. Is there a planned migration / deprecation period, or are the old names officially retired? Are TELEOP_* and CF_TELEOP_* env vars officially retired, or should there be a deprecation shim that reads the old names and warns before this merges?

_{Reviews (1): Last reviewed commit: "Cleanup" | Re-trigger Greptile}

[greptile-apps]

model_copy bypasses type coercion for CLI/env overrides

model_copy(update=...) in Pydantic v2 does not run validators by default, so non-string fields receive raw strings from the CLI or env. For example, setting TRANSPORTS__BROKER__ORDERED=false stores the string "false" — which is truthy — into ordered: bool, and TRANSPORTS__BROKER__HEARTBEAT_HZ=2.0 stores "2.0" into heartbeat_hz: float, which raises TypeError the first time arithmetic is done on it. The fix is to re-validate after the merge, e.g. type(self._config).model_validate({**self._config.model_dump(), **overrides}).

[greptile-apps]

Same model_copy validation gap on WebRTCVideoTransport

The WebRTCVideoTransport.with_config_overrides implementation has the same issue: self._config.model_copy(update=dict(overrides)) won't coerce string values coming from the CLI/env into the declared field types (e.g., ordered: bool, max_retransmits: int | None). It should use model_validate to re-run coercion, consistent with whatever fix is applied to WebRTCTransport.