/ai-skill-steps

Foundational GET all handler for: ai-skill-steps
An AI Skill Step is used inside of a Skill.

It can be one of a few options that are defined at the Global (Sphere=0) level:

Single Shot - Meaning that it is a custom prompt that is given with an input from the user.

HTTPS Call - Meaning that this will make an HTTPS call to a third party system, passing a variable for the headers and body that is configurable by the user.

Skill ID - This Step will call another skill, passing all prior steps as input.

Vector Query - This step will create a Vector Query on the FILES table based on the parameters selected. It requires a prompt and message to be set. It will additionally accept a model default and/or constraints.
ID: NUMBER: This is the Primary Key (Numeric) for the AI Skill Step
LLM_MODEL_DEPLOYMENTS_ID: NUMBER: This (optionally) mandates the LLM deployment that this skill step will use
AI_SKILLS_ID: NUMBER:
AI_SKILL_STEPS_ID: NUMBER: This is the Parent Skill Step if applicableThis is used in branching logic
AI_SKILL_STEP_TYPES_ID: NUMBER: This represents the type of skill step that is being used
STEP_ORDER: NUMBER: This is the numerical Order of the Skill Step
MACHINE_NAME: VARCHAR2(255): This is the Machine Name of this AI Skill Step
TITLE: VARCHAR2(255): This is the Title of this AI Skill Step
DESCRIPTION: CLOB: This is a Description of the AI Skill Step
SYSTEM_MESSAGE: CLOB: This is the SYSTEMMESSAGE that is sent to the LLM if this Skill Step involves a call to LLM
USER_MESSAGE: CLOB: This is the User Message that is sent to the LLM if this step involves an LLM Call
FILE_PARAMETER_FILTERS: JSON: JSON array of any file filter parameters for the Vector searchEach element should include:- Field name- Operator- Valueeg an array with elements: titleeqOperating Agreement
HTTP_HEADERS: JSON: JSON Payload of the HTTP Headers to be set in making this call if an HTTPHTTPS Call
HTTP_BODY: JSON: This is the HTTP Body that is sent in the request if of type HTTPHTTPS request
URL: VARCHAR2(255): This is the HTTP URL for the Skill Step if of type HTTPHTTPS Request
CEPHISSO_PLATFORM_ACCOUNTS_ID: NUMBER: This designates the Sphere that this Skill Step exists in
SYSTEM_GUID: VARCHAR2(36): This is used to track a unique identifier for a system specificaly a System GUID generated by SYSGUID()
UUID: VARCHAR2(36): This is the UUID for this record
METADATA: JSON: This is a JSON metadata field
DATA_LINEAGE: JSON: This field includes any specifications about where the source system is any external identifier time last updated etc
UDFS: JSON: This is used to track the UDFs (User Defined Fields) that might get added
CREATED_ON: DATE: The Date-time that this record was created
CREATED_BY: VARCHAR2(255): The User ID of the user account that created this record
LAST_UPDATED_ON: DATE: This is the Date-time when this record was Last Updated
LAST_UPDATED_BY: VARCHAR2(255): The User ID of the user account who last updated this record
AI_SKILL_ID_TO_EXECUTE_ID: NUMBER: This field is used to indicate if there is a skill to execute as this skill stepThere is a skill Step Type w Machine Name SKILL that will drive the App Logic to determine if it looks at this or not

Provide your bearer token in the

Authorization

header when making requests to protected resources.

Example:

Authorization: Bearer ********************

{ "items": [ { "ai_skills_id": 0, "ai_skill_id_to_execute_id": 0, "ai_skill_steps_id": 0, "ai_skill_step_types_id": 0, "cephisso_platform_accounts_id": 0, "created_by": "string", "created_on": "2019-08-24T14:15:22Z", "data_lineage": null, "description": "string", "file_parameter_filters": null, "http_body": null, "http_headers": null, "id": 0, "last_updated_by": "string", "last_updated_on": "2019-08-24T14:15:22Z", "llm_model_deployments_id": 0, "machine_name": "string", "metadata": null, "step_order": 0, "system_guid": "string", "system_message": "string", "title": "string", "udfs": null, "url": "string", "user_message": "string", "uuid": "string" } ] }

Request

Responses