> ## Documentation Index > Fetch the complete documentation index at: https://docs.uselamina.ai/llms.txt > Use this file to discover all available pages before exploring further. # Refine Run > Iterative refinement: provide natural-language feedback on a completed run's outputs. An LLM identifies which workflow nodes need prompt changes, re-executes them, and returns updated outputs with a summary of what changed. Only works on runs in `completed` status. The refinement happens in-place on the same run — no new run ID is created. **Example:** After generating a product image, send `"Make the background brighter and add more contrast"` to refine the output without re-running the entire workflow. Iterative refinement for completed runs. Provide natural-language feedback and the engine re-executes only the nodes that need to change. The feedback agent uses an LLM to analyze which workflow nodes (image generation, video editing, etc.) need prompt modifications to address the feedback, rewrites only those prompts, and re-executes the affected subgraph. Use this for quality feedback loops: generate content, review it, then refine without starting over. ## OpenAPI ````yaml POST /v1/runs/{runId}/feedback openapi: 3.1.0 info: title: Lamina Apps API version: 1.0.0 description: > Professional creative content production for AI agents. Create brand-aligned images, videos, and designs from natural-language briefs using 30+ specialized AI pipelines. Unlike raw model APIs (DALL-E, Flux, Kling), Lamina automatically applies brand guidelines, selects the optimal multi-model pipeline, scores output quality, and delivers permanent CDN-hosted URLs with composability metadata for chaining. ## Authentication All endpoints require an API key passed via one of these headers: - `x-api-key: lma_your_key` (recommended) - `Authorization: Bearer lma_your_key` API keys are workspace-scoped. Create them in **Settings -> API Keys**. ## Quick Start (One-Call Path) `POST /v1/create` with `{ "brief": "product photo of sneakers", "sync": true }` — returns completed output inline with CDN URL. ## Quick Start (Advanced) 1. `GET /v1/apps` -- browse 30+ specialized creative apps 2. `GET /v1/apps/{appId}` -- inspect input parameters 3. `POST /v1/apps/{appId}/runs?webhook=` -- run with your inputs 4. Receive results via webhook, poll `GET /v1/runs/{runId}`, or stream via `GET /v1/runs/{runId}/stream` ## Rate Limits All `/v1/*` endpoints are currently limited to 100 requests per minute per IP. On `429`, read the `RateLimit-*` and `Retry-After` headers before retrying. ## How Inputs Work When running an app, provide inputs as a JSON object keyed by parameter **name** (from the app details response). - **text** parameters: send a string value (e.g. a prompt or description) - **options** parameters: send the **label** of the option (e.g. `"Caucasian"`), not an internal value - **url** parameters: send a publicly accessible URL to an image or video Every parameter is returned with `required: true`. Parameters with a `default` can safely be omitted -- the app will use the default value. ## Endpoint Groups | Group | Purpose | |-------|---------| | **Apps** | Discover and inspect available content creation apps | | **Runs** | Run workflows, track progress, and retrieve results | | **Assets** | Browse generated images, videos, and text from past executions | | **Intelligence** | Brand context, performance prediction, recommendations, and trends | | **Publishing** | Publish content to connected social channels | | **Content** | One-call agent operations: create, score, brief, batch | | **Account** | Credit balance and rate limit visibility | | **Templates** | Content creation templates and strategies | | **Webhooks** | Webhook signature verification | servers: - url: https://app.uselamina.ai description: Production security: - apiKey: [] tags: - name: Create description: Create content from briefs, run specific apps, and use templates - name: Discover description: Find and inspect available apps and their input schemas - name: Track description: Monitor execution progress, retrieve results, and browse generated assets - name: Intelligence description: >- Brand context, performance prediction, content recommendations, and trend signals - name: Distribute description: Publish content to social channels and transfer assets to CDN - name: Score description: Evaluate content quality, brand alignment, and engagement potential - name: Compose description: >- Turn a script into a narrated, multi-shot video — pick a format (never a model), preview cost, and render - name: Account description: Credit balance, rate limits, and health checks - name: Webhooks description: Webhook signature verification for secure callback handling paths: /v1/runs/{runId}/feedback: post: tags: - Track summary: Refine Run with Feedback description: > Iterative refinement: provide natural-language feedback on a completed run's outputs. An LLM identifies which workflow nodes need prompt changes, re-executes them, and returns updated outputs with a summary of what changed. Only works on runs in `completed` status. The refinement happens in-place on the same run — no new run ID is created. **Example:** After generating a product image, send `"Make the background brighter and add more contrast"` to refine the output without re-running the entire workflow. operationId: refineRun parameters: - name: runId in: path required: true schema: type: string format: uuid description: The run to refine. requestBody: required: true content: application/json: schema: type: object properties: feedback: type: string description: >- Natural-language feedback describing what to change in the outputs. required: - feedback example: feedback: >- Make the background brighter and increase contrast on the product responses: '200': description: Refinement result with updated outputs content: application/json: schema: type: object properties: data: type: object properties: runId: type: string format: uuid status: type: string enum: - completed - failed summary: type: string description: LLM-generated description of what was changed. refinedNodes: type: array items: type: object properties: nodeId: type: string reason: type: string description: Which nodes were re-executed and why. outputs: type: array items: type: object properties: id: type: string label: type: string type: type: string value: type: string description: Full updated output set after refinement. errorMessage: type: string nullable: true example: data: runId: 99ebc7c6-24f3-4fc2-9d53-2e3f49f3bec1 status: completed summary: Brightened the background and increased product contrast refinedNodes: - nodeId: designer-a reason: Background brightness and contrast adjustment outputs: - id: designer-a label: Product Hero type: image value: https://cdn.uselamina.ai/outputs/refined-hero.png errorMessage: null '400': description: Run is not in completed state content: application/json: schema: $ref: '#/components/schemas/Error' '401': $ref: '#/components/responses/Unauthorized' '404': $ref: '#/components/responses/NotFound' components: schemas: Error: type: object description: Standard error response required: - error properties: error: type: string description: Human-readable error message details: type: array items: type: string description: Additional validation error details (present on 400 responses) responses: Unauthorized: description: Missing or invalid API key content: application/json: schema: type: object properties: error: type: string example: error: Invalid API key NotFound: description: Resource not found content: application/json: schema: type: object properties: error: type: string example: error: App not found securitySchemes: apiKey: type: apiKey in: header name: x-api-key description: 'Workspace API key. Prefix: `lma_`. Example: `lma_abc123...`' ````