Skip to main content

Example workflows

These end-to-end workflows show how to combine MCP tools to accomplish common content tasks. Tool names are real; for each tool's exact arguments, read its schema from your client. The server is self-describing.

1. Generate and publish a LinkedIn post​

The most common workflow: brief, generate, review, publish.

1. get_engine_config          // Confirm LinkedIn is enabled
2. create_brief               // Set topic, angle, key points
3. generate_content           // Start generation with briefId
4. get_generation_status      // Poll until status: "done"
5. get_content                // Read the draft
6. send_chat_message          // (Optional) Refine the draft
7. get_connected_channels     // Get LinkedIn connection ID
8. publish_now                // Post immediately

Full example prompt to an AI client connected to Reactor:

"Write a LinkedIn post about why developer tools with opinionated defaults win over flexible frameworks. Use a counterintuitive angle. Keep it under 1300 characters. Publish it to our LinkedIn account."

2. Schedule a week of content​

Build a full week of posts across platforms and schedule them at optimal times.

1. get_engine_config                 // Check which channels are active
2. get_posting_recommendations       // Get best times per platform
3. get_analytics_overview            // See what's already scheduled this week
4. get_calendar                      // Check for gaps in schedule
5. create_brief (x3-5)              // Create briefs for each platform
6. generate_content (each)           // Generate all drafts
7. get_generation_status (each)      // Wait for each to complete
8. schedule_content (each)           // Schedule at recommended times
9. get_calendar                      // Verify week is populated

3. Analyze and replicate top performers​

Use analytics to understand what works and create more of it.

1. get_analytics_overview            // See overall performance period
2. get_top_content                   // Identify top performers
3. get_topic_clusters                // Understand which topics win
4. suggest_brief_from_analytics      // Generate a data-driven brief
5. create_brief                      // Create the brief from the suggestion
6. generate_content                  // Generate
7. get_generation_status             // Wait
8. get_posting_recommendations       // Get best time to post
9. schedule_content                  // Schedule at optimal time

4. Find and act on keyword opportunities​

Use GSC data to identify content gaps and generate posts to close them.

1. get_search_analytics              // Get keyword opportunities
2. create_brief                      // Brief on the top opportunity keyword
   // topic: "Guide to [keyword]"
   // angle: "Optimize for [query] intent"
3. generate_content                  // Generate blog post draft
4. get_generation_status             // Wait
5. publish_now                       // Post to blog (Webflow or a GitHub repo)
6. generate_utm_link                 // Tag the blog URL for tracking

5. Set up automated content cadence​

Configure automation so Reactor keeps publishing even when you are not actively creating.

1. list_automation_rules             // Check what exists
2. create_automation_rule            // Add posting gap rule per platform
   // trigger: posting_gap (3 days) on linkedin
   // action: create_brief (auto-topic)
3. create_automation_rule            // Add top performer follow-up rule
   // trigger: engagement_threshold (5%) on x
   // action: create_brief (expand on insight)
4. list_automation_rules             // Confirm rules are active

Once these rules are in place, Reactor will automatically create new briefs when your publishing cadence lapses or when a post significantly outperforms expectations.

6. Set up a recurring content schedule​

Configure Reactor to auto-generate and publish on a fixed cadence without per-brief intervention.

1. list_schedules              // See what's already running
2. get_posting_recommendations // Best times per platform from analytics
3. create_schedule             // Set the cadence
   // channel_key: "x"
   // platform: "x"
   // frequency: "weekly"
   // day_rules: { "tue": { time: "09:00" }, "thu": { time: "14:00" } }
   // topic: "Product updates and engineering insights"
   // publish_status: "ready"  // auto-publish when generated
4. list_cards                  // Verify placeholder cards appeared on calendar
5. list_schedules              // Confirm schedule is active + next_run_at is set

To pause a schedule without deleting it:

update_schedule    // active: false

To change the cadence (e.g. weekly → biweekly):

update_schedule    // frequency: "biweekly"
                   // confirm_removal: true (if generated posts exist beyond new slots)

7. Upload context and refresh generation​

When you update your product or messaging, refresh the reference documents so future content reflects the change.

1. list_documents                    // See what's currently uploaded
2. upload_document                   // Upload new product spec or messaging doc
   // type: "product_context"
3. create_brief                      // Brief on a feature announcement
4. generate_content                  // Generate -- will incorporate new doc
5. get_generation_status
6. get_content                       // Review the draft
7. schedule_content                  // Schedule for launch day