Skip to main content
Add these nodes with the pipeline builder: pipeline.add(name="...").<node>(...). Each entry lists the node’s configuration parameters. See the Pipeline reference for add, run, and lifecycle methods.

​
add_node

Add a new node to the conversation. 
pipeline.add(name="node").add_node()

​
agent

Agent
Platform docs: Dynamic node: inputs/outputs determined at runtime.
pipeline.add(name="node").agent(use_existing_agent=True, agent=...)
Parameters
​
use_existing_agent
bool
required
​
agent
AcceptsAgent
required
​
model
str
default:"'gpt-5.4'"
​
agent_config
Any
default:"{}"
​
processed_inputs
dict
default:"{}"
​
processed_outputs
dict
default:"{}"
​
provider
str
default:"'openai'"

​
code_execution

Execute code in the language of your choice.
Platform docs: Execute code in the language of your choice.
pipeline.add(name="node").code_execution(code="...")
Parameters
​
code
str
required
​
function_name
str
default:"'main'"
​
input_schema
ListType | list[NameTypeValueAny]
default:"[]"
​
language
int
default:"0"
​
output_schema
ListType | list[NameType]
default:"[]"
​
processed_inputs
dict
default:"{}"
​
processed_outputs
dict
default:"{}"

​
code_interpreter

Execute Python code in a sandboxed environment with file I/O support. Supports spreadsheet analysis, data visualization, and generating artifacts. To read or edit uploaded files, pass them in the files input as an actual array of file references; those files are mounted at /work/inputs/{filename} (curly braces shown for clarity; replace with the exact filename). Continued sessions restore prior inputs and generated files, so work can resume even if Python memory reset. Save output files to /work/outputs/ to have them collected as artifacts; call save_file(…) for durable/versioned user-visible files. Example: files=[“$history.69dd11a49033ae4d8e65ab33”] and pd.read_csv(‘/work/inputs/data.csv’); plt.savefig(‘/work/outputs/chart.png’). You will see stdout and stderr outputs after each execution step. Include an existing session id to keep executing code within the same session. 
pipeline.add(name="node").code_interpreter(code="...")
Parameters
​
code
str
required
​
files
AcceptsFileList
default:"[]"

​
condition

Platform docs: Condition — Specify a series of conditions and execute different paths based on the value of the conditions.
pipeline.add(name="node").condition()
Parameters
​
conditions
list[ConditionGroup]
​
paths
list[str]
​
outputs
dict

​
convert_type

Convert value from source type to target type. 
pipeline.add(name="node").convert_type(source_type="string", value="...")
Parameters
​
target_type
str
default:"'int32'"
The type to convert the value to. One of: bool, float, int32, timestamp
​
source_type
str
required
The type of the value to convert. One of: string
​
value
str
required
The value to convert

​
create_skill

Create a new skill with a name, description, and content 
pipeline.add(name="node").create_skill(content="...", name="...")
Parameters
​
content
str
required
​
name
str
required
​
skill_description
str
default:"''"

​
custom_group

Group multiple nodes together 
pipeline.add(name="node").custom_group()

​
get_skill

Retrieve the full content of a skill by its ID 
pipeline.add(name="node").get_skill(skill=...)
Parameters
​
skill
Any | ListType | list[Any]
required

​
list_objects

List VectorShift objects accessible to the current user 
pipeline.add(name="node").list_objects(object_type="agent")
Parameters
​
include_shared
bool
default:"True"
​
limit
int
default:"0"
​
object_type
str
required
One of: agent, knowledge_base, pipeline, prompt, table, transformation
​
offset
int
default:"0"
​
verbose
bool
default:"False"

​
make

pipeline.add(name="node").make()
Parameters
​
payload
str
default:"''"
​
webhook_url
str
default:"''"

​
merge

Recombine paths created by a condition node. Note: if you are not using a condition node, you shouldn’t use a merge node
Platform docs: Recombine paths created by a condition node. Note: if you are not using a condition node, you shouldn’t use a merge node
pipeline.add(name="node").merge(fields=...)
Parameters
​
function
str
default:"'first'"
The function to apply to the input fields One of: first, join
​
type
str
default:"'string'"
The expected type of the input and output fields
​
fields
ListType | list[Any]
required

​
pipeline

{{collection.pipelines}}
Platform docs: Dynamic node: inputs/outputs determined at runtime.
pipeline.add(name="node").pipeline()
Parameters
​
pipeline
AcceptsPipeline
default:"{}"

​
set_variable

Set a variable to a new value
Platform docs: Set a variable to a new value
pipeline.add(name="node").set_variable(variable_set=..., scope="...", variable_id="...", value="...")
Parameters
​
variable_set
AcceptsVariableSet
required
​
scope
str
required
​
variable_id
str
required
​
value
str
required

​
share_object

Share a VectorShift object with another user 
pipeline.add(name="node").share_object(user_identifier="...")
Parameters
​
object_type
str
default:"'knowledge_base'"
One of: knowledge_base
​
object
AcceptsKnowledgeBase
​
org_name
str
default:"''"
Enter the name of the organization of the user (leave blank if not part of org)
​
user_identifier
str
required
Enter the username or email of the user you want to share with

​
start_flag

Start a conversation
Platform docs: Start a conversation
pipeline.add(name="node").start_flag()

​
time

Outputs the current time (often connected to LLM node) 
pipeline.add(name="node").time()
Parameters
​
delta_time_unit
str
default:"'Seconds'"
One of: days, hours, minutes, seconds, weeks
​
delta_value
int
default:"0"
​
is_positive
str
default:"'+'"
One of: +, -
​
is_positive_delta
bool
default:"True"
​
output_format
str
default:"'MM/DD/YYYY'"
One of: MM-DD-YYYY / HH:MM:SS, MM/DD/YYYY, Timestamp
​
time_node_zone
str
default:"'America/New_York'"
​
use_delta
bool
default:"False"

​
transformation

Use Python code to create a custom node 
pipeline.add(name="node").transformation()

​
wait_node

Pause Pipeline execution for a specified duration 
pipeline.add(name="node").wait_node(wait_time=0, wait_time_type="micros")
Parameters
​
wait_time
int
required
​
wait_time_type
str
required
One of: micros, milliseconds, nanos, seconds