Social Neuron
Developer

Workflow Recipes

These are common multi-step patterns for using Social Neuron's MCP tools. Each workflow can be run by asking Claude in natural language, or by calling the tools programmatically.

Weekly Content Batch

Plan and schedule an entire week of content in one session.

Code
1. get_ideation_context(project_id) -- understand what works 2. generate_content(prompt, type, project) -- create 5 pieces of content 3. quality_check(content, platform) -- validate each piece 4. get_best_posting_times(platform) -- find optimal slots 5. schedule_post(content, platform, time) -- queue everything

Ask Claude: "Plan and schedule 5 posts for this week based on what's been performing well."

Performance Review

Analyze your recent content performance and get actionable next steps.

Code
1. fetch_analytics(days=7) -- raw performance data 2. get_performance_insights(project_id) -- AI analysis 3. get_loop_summary(project_id) -- actionable next steps

Ask Claude: "Review my content performance from the last 7 days and tell me what to do differently."

Cross-Platform Repurposing

Create content once, then adapt it for every platform.

Code
1. generate_content(prompt, platform="youtube") -- primary content 2. adapt_content(content, target="instagram") -- adapt for Instagram 3. adapt_content(content, target="tiktok") -- adapt for TikTok 4. schedule_post(content, platform, time) for each -- distribute everywhere

Ask Claude: "Take my latest YouTube video concept and adapt it for Instagram and TikTok, then schedule all three."

Brand-Aware Autopilot

Set up automated posting that stays true to your brand voice.

Code
1. get_brand_profile(project_id) -- verify brand voice 2. update_autopilot_config(schedule, platforms) -- configure schedule 3. get_autopilot_status() -- verify it's running

Ask Claude: "Set up autopilot to post 3 times per week on YouTube and Instagram using my brand voice."

Comment Engagement

Stay on top of audience interactions across platforms.

Code
1. list_comments(platform, status="pending") -- see new comments 2. moderate_comment(id, action) -- flag spam 3. reply_to_comment(id, reply_text) -- engage audience

Ask Claude: "Show me all pending comments and help me reply to the most important ones."

Full Autonomous Loop

The complete closed-loop content cycle that gets smarter over time.

Code
1. get_loop_summary(project_id) -- assess project state 2. plan_content_week(project_id) -- generate content plan 3. generate_video / generate_image -- create media assets 4. check_status(job_id) -- poll until ready 5. schedule_post(content, platform, time) -- distribute content 6. (wait for analytics collection) -- automatic 7. refresh_platform_analytics() -- trigger manual refresh 8. get_performance_insights(project_id) -- read what worked 9. Loop back to step 1 with new insights

Each iteration produces smarter content as performance data feeds back into the planning step.

Ask Claude: "Run a full content loop -- check what's working, plan new content, create it, and schedule it."

Content Plan with Approval

For teams that need review before publishing.

Code
1. plan_content_week(project_id) -- generate plan 2. quality_check_plan(plan_id) -- validate all posts 3. submit_content_plan_for_approval(plan_id) -- send for review 4. list_plan_approvals(plan_id) -- check status 5. schedule_content_plan(plan_id) -- publish approved posts

Ask Claude: "Create a content plan for next week, run quality checks, and submit it for team approval."

Tips for Better Workflows

Be specific with context. The more context you give Claude about your goals, audience, and brand, the better the output. Instead of "create content," try "create 3 Instagram Reels about sustainable fashion tips for Gen Z."

Chain tools naturally. You do not need to call tools one by one. Describe the full workflow you want and Claude will chain the right tools together automatically.

Use the learning loop. Always start with get_ideation_context or get_loop_summary before creating new content. This ensures your new content is informed by what has performed well in the past.

Last modified on