This page is being phased out as part of our documentation reorganization.
Click this card to be redirected to the updated version with the most current information.
If you notice any discrepancies or areas needing improvement in the new documentation, please use the “Report an issue” button at the bottom of the page.
This article will briefly introduce common structures in plugin development.
When specifying file paths in Manifest or any yaml files, follow these two rules based on file types:
icon
), place them in the _assets
folder under the plugin root directory..py
or .yaml
, use the absolute path within the plugin project.When defining plugins, some data structures can be shared among tools, models, and Endpoints. Here are these shared structures.
I18nObject
is an internationalization structure compliant with IETF BCP 47 standard, currently supporting four languages:
ProviderConfig
is a common provider form structure, applicable to both Tool
and Endpoint
name
(string): Form item namelabel
(I18nObject, required): Follows IETF BCP 47type
(provider_config_type, required): Form typescope
(provider_config_scope): Option range, varies with type
required
(bool): Cannot be emptydefault
(any): Default value, only supports basic types float
int
string
options
(list[provider_config_option]): Options, only used when type is select
helper
(object): Help documentation link label, follows IETF BCP 47url
(string): Help documentation linkplaceholder
(object): Follows IETF BCP 47value
(string, required):valueslabel
(object, required):comply with IETF BCP 47secret-input
(string):Configuration information will be encryptedtext-input
(string):Plain textselect
(string):drop-down boxboolean
(bool):switchgearmodel-selector
(object):Model configuration information, including vendor name, model name, model parameters, etc.app-selector
(object):app idtool-selector
(object):Tool configuration information, including tool vendor, name, parameters, etc.dataset-selector
(string):TBDtype
is model-selector
all
llm
text-embedding
rerank
tts
speech2text
moderation
vision
type
is app-selector
all
chat
workflow
completion
type
is tool-selector
all
plugin
api
workflow
provider
(string): Model provider name including plugin_id, in the format of langgenius/openai/openai
model
(string): Specific model namemodel_type
(enum): Model type enumeration, refer to this documentinputs
(dict): Variables finally input to the nodeoutputs
(dict): Node output resultsprocess_data
(dict): Data generated during node executionprovider_id
(string): Tool provider nametool_name
(string): Tool nametool_description
(string): Tool descriptiontool_configuration
(dict[str, Any]): Tool configuration informationtool_parameters
(dict[str, dict]): Parameters requiring LLM inference
name
(string): Parameter nametype
(string): Parameter typerequired
(bool): Whether requireddescription
(string): Parameter descriptiondefault
(any): Default valueoptions
(list[string]): Available optionsEdit this page | Report an issue