> ## Documentation Index
> Fetch the complete documentation index at: https://docs.superlines.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Automation Workflows

> Build automated AI search visibility pipelines using the Superlines MCP server with AI agents.

The Superlines MCP server isn't just for querying data — it's a building block for automated workflows that connect AI search intelligence to content production, monitoring, and optimization.

## What you can automate

<CardGroup cols={2}>
  <Card title="Prompt Discovery" icon="magnifying-glass">
    Automatically find and add relevant prompts from external data sources.
  </Card>

  <Card title="Competitive Monitoring" icon="chart-bar">
    Surface prompts where you're losing to competitors and trigger content actions.
  </Card>

  <Card title="Content Pipeline" icon="pen-to-square">
    Feed AI search insights directly to content production agents.
  </Card>

  <Card title="Reporting" icon="file-lines">
    Generate automated performance reports and alerts.
  </Card>
</CardGroup>

***

<h2 id="prompt-discovery">
  Prompt discovery workflows
</h2>

### External data source to Superlines

Use AI agents to discover prompts from external sources and add them to tracking automatically.

**Flow:**

1. An AI agent scrapes an external source (SERP results, trending articles, Reddit threads, industry forums)
2. The agent extracts relevant questions and topics
3. The agent calls `list_brands` to identify the target brand
4. The agent calls `add_prompts` to add discovered prompts with appropriate labels

**Example prompt to your AI agent:**

```
Using the Superlines MCP server:
1. List my brands to find the domain ID for "My Brand"
2. Research trending topics in [your industry] using SERP data
3. Generate 10 conversational prompts from those topics
4. Add them to my brand with the label "trend-discovery"
```

### Google Search Console to strategic prompts

Combine GSC data with Superlines analytics to find high-value prompts:

**Flow:**

1. Connect [Google Search Console](/integrations/google-search-console) to import queries
2. Use an AI agent to call `get_query_data` and identify high-impression queries
3. Cross-reference with `analyze_metrics` to find queries where visibility is low
4. Prioritize these as strategic prompts with `update_prompt_labels`

***

## Competitive intelligence workflows

### Find prompts where you're losing

Automatically identify fan-out queries where competitors outrank you.

**Example prompt to your AI agent:**

```
Using the Superlines MCP server for brand "My Brand":
1. Call get_fanout_query_insights to see fan-out query performance
2. Identify queries where my domain ranks poorly or isn't present
3. Call get_competitive_gap to see which competitors lead on those topics
4. For the top 5 gaps, call webpage_audit on the competitor's cited page
5. Summarize: what content am I missing, and what should I create?
```

### Audit competitor content automatically

When a competitor is winning on a specific prompt, reverse-engineer their strategy:

**Flow:**

1. Call `get_top_cited_url_per_prompt` to find which URLs are winning
2. Call `webpage_audit` on those competitor URLs to get detailed content analysis
3. Call `webpage_audit` on your own competing pages for comparison
4. Generate a content brief based on the gaps

***

## Content production pipeline

### Agent-to-agent content collaboration

The most powerful workflow connects Superlines data directly to a content production agent. This creates a closed-loop system: **discover opportunities → create content → monitor results**.

**Architecture:**

```
┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
│  Superlines MCP │────▶│  Orchestrator    │────▶│  Content Agent  │
│  (Intelligence) │     │  Agent           │     │  (Production)   │
│                 │◀────│                  │◀────│                 │
└─────────────────┘     └──────────────────┘     └─────────────────┘
```

**Step-by-step flow:**

<Steps>
  <Step title="Identify opportunities">
    The orchestrator calls `find_content_opportunities` to discover topics with high query volume but low brand visibility.
  </Step>

  <Step title="Analyze competitive landscape">
    For each opportunity, call `get_fanout_query_insights` and `get_top_cited_url_per_prompt` to understand what content currently wins.
  </Step>

  <Step title="Audit top-performing content">
    Call `webpage_audit` and `webpage_analyze_content` on the winning URLs to extract what makes them successful.
  </Step>

  <Step title="Generate content brief">
    The orchestrator synthesizes the data into a content brief: target prompts, fan-out queries to optimize for, competitive gaps to exploit, and content structure recommendations.
  </Step>

  <Step title="Content agent produces draft">
    A content production agent (Claude, GPT, or a custom agent) creates content based on the brief. This could feed directly into a CMS publication queue.
  </Step>

  <Step title="Monitor results">
    After publishing, use `get_weekly_performance` and `get_period_comparison` to track whether the new content improves visibility for the target prompts.
  </Step>
</Steps>

***

## Reporting and monitoring workflows

### Automated weekly reports

**Example prompt to your AI agent:**

```
Using the Superlines MCP server:
1. Call get_weekly_performance for brand "My Brand" over the last 4 weeks
2. Call get_period_comparison to compare this month vs last month
3. Call get_best_performing_prompt to find top 5 prompts
4. Call find_content_opportunities for the top 3 gaps
5. Format a weekly executive summary with key trends and recommended actions
```

### Strategic action plans

**Example prompt to your AI agent:**

```
Using Superlines MCP:
1. Call generate_strategic_action_plan with focusArea "all" for "My Brand"
2. For each recommended action, call the relevant analysis tool to gather supporting data
3. Produce a prioritized action plan with specific content recommendations
```

***

## Prompt management at scale

The MCP server makes prompt management efficient for large portfolios:

| Task                           | MCP approach                                                                          |
| ------------------------------ | ------------------------------------------------------------------------------------- |
| **List current prompts**       | Call `list_prompts` to see all tracked prompts with status and quota usage            |
| **Add prompts in bulk**        | Call `add_prompts` with up to 50 prompts per call, with labels                        |
| **Pause seasonal prompts**     | Call `pause_prompts` with a label filter (e.g., `summer-campaign`)                    |
| **Reactivate after season**    | Call `unpause_prompts` with the same label filter                                     |
| **Reorganize labels**          | Call `update_prompt_labels` to reclassify prompts                                     |
| **Remove stale prompts**       | Call `delete_prompts` with a label or ID list to permanently remove low-value prompts |
| **Discover + add in one flow** | Chain external data scraping with `add_prompts` in a single agent workflow            |

***

## Best practices for MCP automation

<AccordionGroup>
  <Accordion title="Start simple, then automate">
    Run workflows manually with your AI assistant first. Once you understand the flow and validate the outputs, then build automated pipelines.
  </Accordion>

  <Accordion title="Use labels for workflow tracking">
    Label prompts added by automated workflows (e.g., `mcp-auto`, `trend-discovery`) so you can measure their effectiveness separately from manually curated prompts.
  </Accordion>

  <Accordion title="Combine MCP tools in sequence">
    The most powerful workflows chain multiple MCP tools. For example: `list_brands` → `get_fanout_query_insights` → `webpage_audit` → `generate_strategic_action_plan`.
  </Accordion>

  <Accordion title="Keep a human in the loop for content publishing">
    Let agents generate content briefs and drafts, but have a human review before publishing. AI search visibility is a strategic asset — quality matters more than speed.
  </Accordion>

  <Accordion title="Monitor automated prompts">
    Regularly review prompts added by automated workflows. Pause or remove any that aren't generating valuable insights after 2-4 weeks of tracking.
  </Accordion>
</AccordionGroup>

***

## Getting started

<CardGroup cols={2}>
  <Card title="MCP Setup" icon="plug" href="/mcp/setup">
    Connect the MCP server to your AI assistant.
  </Card>

  <Card title="Available Tools" icon="wrench" href="/mcp/tools">
    Full reference of all 30 MCP tools.
  </Card>
</CardGroup>