> ## Documentation Index > Fetch the complete documentation index at: https://docs.puffle.ai/llms.txt > Use this file to discover all available pages before exploring further. # Add Node > Insert a new node into a draft campaign's sequence. Also the home page for moving and deleting nodes on the same collection. **CLI:** ```bash theme={null} puffle campaign sequence node add --id --type puffle campaign sequence node add --id --type --position --subject --body --signature --include-message --is-reply --delay-days --delay-hours --delay-minutes ``` ## Overview Inserts a new node into the campaign sequence. **Draft campaigns only** — every mutation on this collection is gated on `status === "draft"`. This collection exposes three verbs: | Verb | Operation | Purpose | | -------- | -------------------- | ----------------------------------------------------------------------------------------------------- | | `POST` | `addSequenceNode` | Insert a new node (this page). Omit `position` to drop it right before the terminal `end` node. | | `PATCH` | `moveSequenceNode` | Reorder an existing node — see [Move a sequence node](/api-reference/campaigns/update-sequence-node). | | `DELETE` | `deleteSequenceNode` | Remove a node — see [Delete a sequence node](/api-reference/campaigns/delete-sequence-node). | To edit a node's content (subject, body, delay values, AI purpose), use [`updateSequenceNode`](/api-reference/campaigns/update-sequence-node) against the singular `/sequence/nodes/{nodeId}` route. Type is immutable after creation — to change it, delete and re-add. Every mutation on this collection is backed by an atomic Postgres RPC with row-level locking, so concurrent adds, moves, and deletes can't corrupt the `position` ordering. ## AI agent notes **Typical build order.** 1. `POST /api/campaigns` to create a draft. 2. Loop over your desired nodes and `POST /api/campaigns/{id}/sequence/nodes` for each — omit `position` and every node lands just before the auto-inserted terminal `end`. 3. When done, `POST /api/campaigns/{id}/launch`. **Channel validation.** The parent campaign's `type` determines which node kinds are legal: * `linkedin` campaigns accept `connection_request`, `message`, `delay`, `ai`, `end`. * `email` campaigns accept `email`, `delay`, `ai`, `end`. **Read-back fallback.** The response is normally `{ sequence_nodes: [...] }` — the full list, already ordered. On the rare occasion the post-mutation read-back fails, the server returns `{ mutated: true, node_id }`. In that case refetch via `GET /api/campaigns/{id}`. **Position omitted = before the end.** Don't set `position` unless you specifically need to insert somewhere other than the natural end-of-sequence slot. The server handles ordering. **AI nodes.** Adding `{ type: "ai" }` creates the base row; populate `purpose` and `output_type` with a follow-up `PUT /api/campaigns/{id}/sequence/nodes/{nodeId}` call. ## OpenAPI ````yaml post /api/campaigns/{id}/sequence/nodes openapi: 3.0.3 info: title: Puffle API version: 1.0.0 description: >- HTTP API for the Puffle GTM platform - Feed and Lead Finder search preparation, campaign-backed Outbound execution, Puffle Agent runs, sender management, and Unibox reply workflows. Designed for operation by both humans and autonomous AI agents. Agents should start at the [Agent Playbook](/guides/agent-playbook) which prescribes a workspace-context check (`GET /api/context`) and core user journeys with exact call ordering. contact: name: Puffle Support url: https://puffle.ai servers: - url: https://app.puffle.ai description: Production security: - bearerAuth: [] tags: - name: Campaigns description: >- Campaign-backed Outbound execution. These endpoints manage draft, launching, active, paused, and completed Outbound runs with sequence nodes, contacts, and sender accounts. - name: Accounts description: Signed-in user account, settings, and workspace context endpoints. - name: Billing description: Billing customer portal and account billing endpoints. - name: Senders description: >- Connected sender accounts, email inboxes, sending domains, DNS verification, warmup, and sender capacity. - name: Socials description: >- Read connected LinkedIn and X accounts, review current and archived posts, and reconcile publication status. Creating, editing, scheduling, publishing, OAuth connection, analytics, and mentions remain dashboard-managed workflows and are not part of the public Bearer-token API. paths: /api/campaigns/{id}/sequence/nodes: post: tags: - Campaigns summary: Add Node description: >- Insert a new node into the campaign sequence. Draft campaigns only. If `position` is omitted, the node is placed immediately before the terminal `end` node. operationId: addSequenceNode parameters: - name: id in: path required: true schema: type: string description: Campaign UUID requestBody: required: true content: application/json: schema: type: object properties: type: type: string enum: - connection_request - message - email - delay - ai - end position: type: integer minimum: 0 maximum: 9007199254740991 nullable: true subject: type: string nullable: true body: type: string nullable: true signature: type: string nullable: true include_message: type: boolean nullable: true is_reply: type: boolean nullable: true delay_days: type: integer minimum: 0 maximum: 9007199254740991 nullable: true delay_hours: type: integer minimum: 0 maximum: 9007199254740991 nullable: true delay_minutes: type: integer minimum: 0 maximum: 9007199254740991 nullable: true required: - type additionalProperties: {} responses: '200': description: >- Node inserted. Returns the fresh ordered list, or a mutation fallback when read-back fails. content: application/json: schema: anyOf: - type: object properties: sequence_nodes: type: array items: type: object properties: id: type: string description: String value with endpoint-specific validation. campaign_id: type: string description: String value with endpoint-specific validation. position: type: integer minimum: 0 maximum: 9007199254740991 type: type: string enum: - connection_request - message - email - delay - ai - end subject: type: string nullable: true body: type: string nullable: true body_html: type: string nullable: true signature: type: string nullable: true from_name: type: string nullable: true reply_to: type: string nullable: true cc: type: string nullable: true bcc: type: string nullable: true attachments: type: array items: {} nullable: true include_message: type: boolean nullable: true is_reply: type: boolean nullable: true delay_days: type: integer minimum: 0 maximum: 9007199254740991 nullable: true delay_hours: type: integer minimum: 0 maximum: 9007199254740991 nullable: true delay_minutes: type: integer minimum: 0 maximum: 9007199254740991 nullable: true ai_node: type: object properties: output_type: type: string enum: - email - message required: - output_type additionalProperties: {} nullable: true required: - id - campaign_id - position - type additionalProperties: false required: - sequence_nodes additionalProperties: false - type: object properties: mutated: type: boolean enum: - true node_id: type: string description: String value with endpoint-specific validation. required: - mutated additionalProperties: false '400': description: Validation failure, malformed RPC input, or non-draft campaign. content: application/json: schema: type: object properties: error: type: string required: - error additionalProperties: false '401': description: Missing or invalid authentication. content: application/json: schema: type: object properties: error: type: string required: - error additionalProperties: false '404': description: Campaign doesn't exist or isn't owned by the caller. content: application/json: schema: type: object properties: error: type: string required: - error additionalProperties: false '500': description: Unexpected server error. content: application/json: schema: type: object properties: error: type: string required: - error additionalProperties: false components: securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: pk_live_... ````