Build and use computer vision models fast with Roboflow. Train, deploy, and run inference on object detection, classification, and segmentation models.
Build and use computer vision models fast with Roboflow. Train, deploy, and run inference on object detection, classification, and segmentation models. On Nagent, Roboflow is exposed as a fully-configurable artificial intelligence integration that any agent can call — 7 actions, and API key authentication. No code is required to wire Roboflow into your workflow — connect it once via the External Integrations panel and reuse it across every agent you build.
Agent builders use Roboflow to automate the kinds of tasks artificial intelligence teams previously handled manually. Concrete examples — each one is a single agent step in Nagent — include:
Every action and trigger is paired with a structured input/output schema (visible in the sections below), so when you wire Roboflow into Helix — our agentic agent builder — the editor knows exactly what each step expects and produces. Configure once, deploy anywhere across your Nagent agents.
Every operation an agent can call against Roboflow, with input parameters and output schema. Drop these into any step of an agent built in Helix.
ROBOFLOW_DESCRIBE_WORKFLOW_INTERFACETool to describe the interface of a workflow specification. Use when you need to understand the inputs, outputs, and types for a Roboflow workflow definition.
Input parameters
Roboflow API Key that will be passed to the model during initialization for artifact retrieval
Workflow specification object containing version, inputs, steps, and outputs. Must include: version (string), inputs (list of input definitions), steps (list of step definitions), and outputs (list of output definitions).
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
ROBOFLOW_GET_EXECUTION_ENGINE_VERSIONSTool to retrieve available Execution Engine versions from Roboflow workflows API. Use when you need to check which execution engine versions are supported for workflow processing.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
ROBOFLOW_GET_SERVER_INFOTool to retrieve information about the Roboflow inference server. Use when you need to check the server version, name, or unique identifier.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
ROBOFLOW_GET_SERVER_METRICSTool to retrieve Prometheus metrics from the Roboflow inference server. Use when monitoring server performance, tracking inference statistics, or debugging server health issues.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
ROBOFLOW_GET_WORKFLOW_SCHEMATool to fetch the workflows block schema from Roboflow. Use when you need to retrieve the complete schema definition for workflow blocks.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
ROBOFLOW_RUN_WORKFLOWTool to run a workflow specification with provided inputs. Use when you need to execute a custom Roboflow workflow pipeline for image processing, model inference, or other computer vision tasks.
Input parameters
Dictionary containing runtime values for each parameter defined as an input in the workflow specification. Keys must match the 'name' fields defined in specification.inputs. Values can be of various types depending on the input type (e.g., for InferenceImage input, provide object with 'type' and 'value' keys where type can be 'url', 'base64', or 'numpy').
Roboflow API Key passed to the model during initialization for artifact retrieval. If not provided, uses the authentication from metadata headers.
Reserved flag used internally by Roboflow to distinguish between preview and non-preview runs.
Optional identifier of the workflow. Used when running a saved workflow definition from Roboflow platform.
Workflow specification object defining version, inputs, steps, and outputs. Must include 'version' (string), 'inputs' (array of input definitions), 'steps' (array of processing steps), and 'outputs' (array of output definitions). Each input must have 'type' and 'name'. Each step must have 'type' and 'name' with additional configuration. Each output must have 'type', 'name', and 'selector'.
List of field names to exclude from the response among those defined in workflow specification outputs.
Flag to request workflow run profiling. Enables workflow profiler only when server settings allow profiling traces to be exported to clients. Only applies for workflow definitions saved on Roboflow platform.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
ROBOFLOW_VALIDATE_WORKFLOWTool to validate a Roboflow workflow specification before execution. Use this to check if your workflow definition is syntactically correct and properly structured.
Input parameters
List of workflow steps that process the inputs and produce intermediate results.
List of workflow inputs that define the data flowing into the workflow.
Roboflow API Key that will be passed to the model during initialization for artifact retrieval. If not provided, uses the authentication from metadata.
List of workflow outputs that extract final results from the steps.
Version of the workflow specification format.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
No publicly available marketplace agent is found using this tool yet. There are 50 agents privately built on Nagent that already use Roboflow.
Build on Nagent
Connect Roboflow to any Nagent agent in minutes — no API key management, no boilerplate. Just configure and deploy.
The five questions agent builders ask before adopting a new integration.
Open the External Integrations panel inside Nagent (app.nagent.ai/externalIntegration), find Roboflow, and click "Connect Now." You'll authenticate with an API key — Nagent handles credential storage and refresh automatically. Once connected, Roboflow is available to any agent in your workspace.
No. Nagent provides no-code integration for every tool. Once Roboflow is connected, you configure its 7 actions directly in the agent builder UI — no API calls, no boilerplate, no schema management.
Helix — Nagent's agentic agent builder — lets you drop Roboflow steps into any workflow visually. Pick an action (e.g., one of those listed above), fill in the inputs (Helix knows the required vs. optional schema for each parameter), and connect it to upstream/downstream steps. Triggers run as the entry point of an agent, so when a Roboflow event fires, the agent kicks off automatically.
Every Roboflow action and trigger ships with a fully-typed schema — input parameters with name, type, required flag, and description, plus the output payload shape. The schemas are documented in the sections above. Helix uses these schemas to validate your configuration at build time and to type-check the data flowing between steps.
Yes. While Roboflow ships with 7 pre-built artificial intelligence actions, you can layer custom logic around them inside Helix — pre/post-processing steps, conditional branches, retries, or stitching Roboflow together with other connected tools. For deeper customization, talk to our team about Nagent's Agentic AI Lab — forward-deployed engineers who build Roboflow-based workflows tailored to your business.