Built-in templates¶

A template creates a ready-made multi-Orchestrator setup in one call: a top-level Orchestrator plus two sub-Orchestrators, already linked agent_tool-style and with their built-in tools already turned on. This page covers listing the catalog and instantiating one. The result is a normal Orchestrator object, see Orchestrators for every field it exposes and how to edit it afterward.

Gated behind a deployment setting, admin session token required

See the two notes at the top of Orchestrators: every endpoint here only exists when your deployment has Orchestrators turned on, and every endpoint here authenticates with an admin session token.

The catalog¶

GET /api/v1/orchestrator-templates

Requires: Viewer and above.

Returns every built-in template this AiFlow version ships.

Field	Type	Notes
`key`	string	The template's identifier, used as the `{key}` path segment when instantiating it.
`name`	string	A human-readable display name.
`description`	string	A short summary of what the resulting Orchestrator does.
`sub_templates`	array	The sub-Orchestrators this template creates, see below.
`sub_templates[].key`	string	The sub-Orchestrator's identifier within this template.
`sub_templates[].name`	string	A human-readable display name for the sub-Orchestrator.
`sub_templates[].description`	string	A short summary of the sub-Orchestrator's role.
`sub_templates[].tool_names`	array of strings	The built-in tools this sub-Orchestrator has turned on, matching the `name` values from Orchestrator tools.

cURLPythonTypeScript

curl https://api.your-domain.com/api/v1/orchestrator-templates \
  -H "Authorization: Bearer $ADMIN_TOKEN"

import httpx

response = httpx.get(
    "https://api.your-domain.com/api/v1/orchestrator-templates",
    headers={"Authorization": f"Bearer {admin_token}"},
)
response.raise_for_status()
catalog = response.json()

const response = await fetch("https://api.your-domain.com/api/v1/orchestrator-templates", {
  headers: { Authorization: `Bearer ${adminToken}` },
});
const catalog = await response.json();

The three built-in templates¶

Each is adapted from a real Google ADK sample agent, narrowed to what needs nothing beyond Gemini or Vertex AI and never pauses mid-run waiting on a human decision.

`key`	What it does	Sub-Orchestrators
`content_claim_auditor`	Fact-checks a piece of text against current web sources and returns a corrected version with a changelog.	`critic` (verifies each claim with `google_search`), `reviser` (rewrites the text using the critic's findings)
`research_report_assistant`	Researches a topic and returns a structured, sourced report.	`researcher` (gathers current information with `google_search`), `synthesizer` (turns raw findings into a report)
`marketing_campaign_assistant`	Proposes a brand name and writes launch copy for a product or business.	`brand_namer` (proposes and checks candidate names with `google_search`), `copywriter` (writes the tagline, website copy, and social posts)

Instantiating a template¶

POST /api/v1/orchestrator-templates/{key}/instantiate

Requires: Admin and above, the same tier creating an Orchestrator by hand requires.

Field	Type	Required	Notes
`slug`	string	Yes	1 to 60 characters, same pattern as a plain Orchestrator's `slug`. Capped shorter to leave room for the longest sub-Orchestrator suffix a template derives from it.
`name`	string	Yes	1 to 120 characters. Used as-is for the top-level Orchestrator; each sub-Orchestrator's own name becomes `"{name}: {sub-Orchestrator name}"`.

{key} must be one of the key values from the catalog above; anything else gets a 404 with {"detail": "Unknown template '<key>'"}. Each sub-Orchestrator's slug is derived automatically, {slug}-{sub_template_key}, hyphenated (for example, instantiating content_claim_auditor with slug: "my-auditor" creates my-auditor, my-auditor-critic, and my-auditor-reviser). If slug, or any slug it would derive, is already in use, nothing is created and the request returns a 409:

{ "detail": "Slug(s) already in use: my-auditor-critic" }

cURLPythonTypeScript

curl -X POST https://api.your-domain.com/api/v1/orchestrator-templates/content_claim_auditor/instantiate \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $ADMIN_TOKEN" \
  -d '{"slug": "site-copy-auditor", "name": "Site Copy Auditor"}'

import httpx

response = httpx.post(
    "https://api.your-domain.com/api/v1/orchestrator-templates/content_claim_auditor/instantiate",
    headers={"Authorization": f"Bearer {admin_token}"},
    json={"slug": "site-copy-auditor", "name": "Site Copy Auditor"},
)
response.raise_for_status()
orchestrator = response.json()

const response = await fetch(
  "https://api.your-domain.com/api/v1/orchestrator-templates/content_claim_auditor/instantiate",
  {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      Authorization: `Bearer ${adminToken}`,
    },
    body: JSON.stringify({ slug: "site-copy-auditor", name: "Site Copy Auditor" }),
  },
);
const orchestrator = await response.json();

Returns 201 Created with the top-level Orchestrator object, the same shape Creating an Orchestrator returns. Its two sub-Orchestrators are created alongside it, already linked underneath it and, where the template calls for it, already tooled: fetch them from Listing Orchestrators or Sub-Orchestrator links if you need their own ids.

Nothing about a template-created row is special¶

A template only runs once, at instantiation time. Afterward, the resulting Orchestrator and its sub-Orchestrators are ordinary rows: edit an instruction, change a model, add or remove a sub-Orchestrator link, enable another tool, or disable one entirely through the same endpoints Orchestrators, Orchestrator links, and Orchestrator tools already document. There's no stored link back to the template that created them, and no re-sync or update-from-template mechanism: a template is a starting point, not an ongoing relationship.