Skip to main content
POST
Clone Template

Authorizations

X-API-Key
string
header
required

API key for authentication.

Path Parameters

template_id
string
required

Body

application/json

Optional overrides when cloning a workflow template.

content_filter
ContentFilter · object | null

Composable content filter using logical operators.

description
string | null
expected_input
Expected Input · object | null

Inputs the template expects. Each key is an abstract snake_case NAME, not a value. Reference a key as {{ key }} in the prompt to substitute its value inline; an unreferenced key is still provided to the model as attached context.

expected_output
string | null
Minimum string length: 1
name
string | null
prompt
string | null
ranking_parameters
RankingParameters · object | null

Parameters that influence how search results are ranked.

research_plan
Plan · object | null

A structured research plan with ordered steps.

tools_configs
Tools Configs · object | null

Per-tool configuration stored with the template, keyed by tool id. Two key families are supported: python_code_execution (toggles sandboxed Python and chart rendering for every run) and connector tool ids in the <connector>__<tool> form (e.g. "factset__FactSet_Fundamentals": {"enabled": true}). Search settings belong in content_filter, ranking_parameters, or the execution request — not here. See GET /v1/tools for each connector tool's config schema.

Example:

Response

Successful Response

A complete workflow template with all metadata.

created_at
string<date-time>
required

When the template was created.

id
string
required

Unique template identifier.

is_community
boolean
required

Whether this is a community-shared template.

name
string
required

Template display name.

prompt
string
required

The research prompt with optional placeholders.

updated_at
string<date-time>
required

When the template was last modified.

content_filter
ContentFilter · object | null

Content filters applied to all searches.

description
string | null

Brief description of what this template does.

expected_input
Expected Input · object | null

Inputs keyed by abstract snake_case NAME. Reference a key as {{ key }} in the prompt to substitute its value inline; an unreferenced key is still provided to the model as attached context.

expected_output
string | null

Optional guidance for the final answer's tone, structure, and format.

Minimum string length: 1
ranking_parameters
RankingParameters · object | null

Search result ranking configuration.

research_plan
Plan · object | null

Pre-defined research plan.

tools_configs
Tools Configs · object | null

Per-tool configuration stored with the template, keyed by tool id. Two key families are supported: python_code_execution (toggles sandboxed Python and chart rendering for every run) and connector tool ids in the <connector>__<tool> form (e.g. "factset__FactSet_Fundamentals": {"enabled": true}). Search settings belong in content_filter, ranking_parameters, or the execution request — not here. See GET /v1/tools for each connector tool's config schema.

Example: