Skip to content
Trading bot
Tentacles development
Create a tentacle package

Tentacle package development

Tentacle packages

This page covers tentacle package creation.

A tentacle package is a python module that contains one or multiple tentacles of the same type.

The tentacle package folder

A tentacle package is defined by a folder located at :


TP is for tentacle package

  • YOUR_TP_CATEGORY can be Backtesting, Evaluator, Services or Trading

  • YOUR_TP_SUB_CATEGORY should be a sub category of YOUR_TP_CATEGORY in the existing tentacle architecture

  • YOUR_TENTACLE_PACKAGE_NAME is the name of your tentacle package, shouldn’t use an existing tentacle package name

Description file

A tentacle package contains metadata described in the metadata.json file. This file is used to properly install the tentacle and should be carefully written. It’s located at the root path of the tentacle package :


A tentacle package metadata.json contains :

  "version": "YOUR_TP_VERSION",
  "origin_package": "YOUR_TP_ORIGIN_PACKAGE",
  "tentacles": ["YOUR_TP_TENTACLE_1", "YOUR_TP_TENTACLE_2"],
  "tentacles-requirements": ["YOUR_TP_TP_REQUIREMENT_1", "YOUR_TP_TP_REQUIREMENT_2"]
  • YOUR_TP_VERSION is your tentacle package version

  • YOUR_TP_ORIGIN_PACKAGE is the author or the origin repository of the tentacle package

  • YOUR_TP_TENTACLE_1 and YOUR_TP_TENTACLE_2 are name of the tentacle classes contained in your tentacle package tentacles (1 or more).

  • YOUR_TP_TP_REQUIREMENT_1 and YOUR_TP_TP_REQUIREMENT_2 are the names of required tentacle packages to have installed to run your tentacle package (0 or more)

YOUR_TP_TENTACLE_X should match python classes to be exposed in the tentacle

Example DailyTradingMode/metadata.json :

  "version": "1.2.0",
  "origin_package": "OctoBot-Default-Tentacles",
  "tentacles": ["DailyTradingMode"],
  "tentacles-requirements": ["mixed_strategies_evaluator"]

Tentacle modules

Tentacle python modules should be placed at the root path of the tentacle package. There can 1 or more modules per package.

Example with momentum_evaluator : The main python module that contains multiple tentacles is located at


Every tentacle classes should be imported in the package root file.

Example with momentum_evaluator’s :

from .momentum import RSIMomentumEvaluator, ADXMomentumEvaluator, RSIWeightMomentumEvaluator, 
BBMomentumEvaluator, MACDMomentumEvaluator, KlingerOscillatorMomentumEvaluator, 


A tentacle package can contain tentacle configurations. Tentacles configuration consists in 2 parts:

  1. User inputs: the description of each configuration parameter, in the init_user_inputs tentacle method
  2. Default configuration: the default configuration values, as json, in the config folder

User inputs

On OctoBot’s web interface, tentacle configuration settings are generated using their definition as User inputs as well as their current values.

For a configuration parameter to show on the configuration interface, it has to be defined as user input. Any value contained in the json configuration file can be used by the tentacle but only the ones associated to user inputs will be visible to the user.

def init_user_inputs(self, inputs: dict) -> None:
   self.period_length = self.UI.user_input(
        "period_length", enums.UserInputTypes.INT, 14, inputs, 
        min_val=1, title="EMA period length."
    self.min_trigger_value = self.UI.user_input(
        "sleep_delay", enums.UserInputTypes.FLOAT, 0.5, inputs, 
        min_val=0.34, max_val=0.75, title="Threshold above which to trigger a signal."
    self.send_notification = self.UI.user_input(
        "send_notification", enums.UserInputTypes.BOOLEAN, True, inputs, 
        title="When enabled, send telegram notification on signal."

The full definition of user inputs can be found here.

If you are unsure how to use user inputs, have a look at the existing tentacles user inputs.

Tentacles configuration are displayed using the json-editor library. User inputs are converted into json schemas that are then passed to the editor alongside their current configuraiton values.

  • The editor_options argument allows to set json-editor specific options such as the disable_array_add option (editor_options={"disable_array_add": True}).
  • The other_schema_values argument allows to set json schema specific parameters such as the minItems or uniqueItems for arrays (other_schema_values={"minItems": 1, "uniqueItems": True}).

Default configuration

Values for an tentacle default configuration are located in the config/ folder at :


Each tentacles config file should be named with the exact case and name as the associated tentacle class. Below an example for MyAwesomeTentacle :


Once a tentacle configuration has been edited, a local copy of this json configuration file is added to your profile where local changes are saved.


Tentacle resources are located in the resources folder of your tentacle package.

Each tentacles documentation should be created for in resources/, resources/ (the file name should match the tentacle class name).

A tentacle package can also contain many resources that can be binary files, images…

Example DailyTradingMode/resources/ :

DailyTradingMode is a **low risk versatile trading mode** that reacts only the its state changes to
a state that is different from the previous one and that is not NEUTRAL.

When triggered for a given symbol, it will cancel previously created (and unfilled) orders
and create new ones according to its new state.

DailyTradingMode will consider every compatible strategy and average their evaluation to create
each state.

You can use markdown to format a tentacle documentation.


Tentacle should be tested. Tests file are usually located in the tests folder of the tentacle package.

Installing and sharing tentacles

Follow the tentacles installation guide to install or share your custom tentacle package.