Hooklayer for n8n
In your n8n workflow, add the "MCP Client" node and configure Hooklayer's URL (https://hooklayer.dev/api/mcp) + Bearer hl_live_ key. The 7 Hooklayer tools (analyze_account, score_hook, viral_remix, trend_pulse, find_viral_template, match_voice, predict_virality) appear in the node's tool dropdown and chain like any other n8n step.
What works in n8n
- All 7 Hooklayer tools as MCP Client node actions
- Bearer auth via header injection
- Chains with n8n's 400+ existing nodes
- Credentials stored in n8n vault
- recommended_chain readable by downstream nodes
Setup (60s)
Config file: n8n MCP Client node settings
# Inside n8n's MCP Client node:
URL: https://hooklayer.dev/api/mcp
Transport: HTTP (Streamable / JSON-RPC 2.0)
Authentication: Header Auth (n8n credential)
# Create credential of type "Header Auth":
Name: Authorization
Value: Bearer hl_live_...
# Once connected, the node's "Tool" dropdown lists:
- analyze_account
- score_hook
- viral_remix
- trend_pulse
- find_viral_template
- match_voice
- predict_virality- 1Get your API key
Sign up at hooklayer.dev/auth/signup for an hl_live_ key. Copy it.
- 2Create an n8n Header Auth credential
In n8n: Settings → Credentials → Create New → Header Auth. Name: "Hooklayer Bearer". Value: "Bearer hl_live_..." (with your real key).
- 3Add the MCP Client node
In your workflow, add the MCP Client node. Configure it as remote HTTP MCP server. Point at https://hooklayer.dev/api/mcp.
- 4Pick a tool from the dropdown
The node's "Tool" dropdown lists Hooklayer's 7 tools. Pick one, configure its inputs (or pass from upstream nodes), and run.
Example prompts for n8n
Paste any of these to see Hooklayer respond live.
Detect viral TikTok videos and email yourself the analysis
# n8n workflow:
# 1. Cron trigger (daily 9am)
# 2. HTTP Request → TikTok trending feed
# 3. Filter → top 5 by views
# 4. MCP Client (Hooklayer): tool=analyze_account, params={handle: $json.username}
# 5. Email node → send summary to your inboxExpected: Daily-cadence pipeline. Auto-analyzes top trending creators each morning. Each analyze_account call costs 5 credits — caching keeps daily cost predictable.
Score every TikTok hook saved to a Google Sheet
# n8n workflow:
# 1. Google Sheets trigger (new row)
# 2. MCP Client (Hooklayer): tool=score_hook, params={text: $json.hook, platform: "tiktok"}
# 3. Update sheet with score, percentile, pattern_match, would_fail_becauseExpected: Per-hook scoring at scale. Useful for content teams reviewing dozens of drafts/week. 1 credit per call; 24-hour determinism cache means re-scoring is free within the window.
Build a 3-step viral pipeline: analyze → remix → predict
# n8n workflow:
# 1. Manual trigger (input: TikTok URL + your topic)
# 2. MCP Client (Hooklayer): tool=analyze_account, params from input
# 3. MCP Client (Hooklayer): tool=viral_remix, params from analyze response.recommended_chain[2]
# 4. MCP Client (Hooklayer): tool=predict_virality, params={script: $json.fresh_script.scenes.map(s => s.voiceover).join(" ")}
# 5. If predict.virality_score >= 70, save to Notion; else loop back to step 3Expected: Automated content quality gate. Cost: 5 + 3 + 2 = 10 credits per pipeline run. Quality threshold (70) keeps slop out of your content calendar.
Frequently asked
Which n8n version do I need?
n8n 1.55.0+ for the MCP Client node (released 2026-Q1). Self-hosted users: docker pull n8nio/n8n:latest. Cloud users: it's in the default node set.
Can I use Hooklayer with the n8n MCP Server node (not Client)?
No — those are different things. MCP Server node EXPOSES n8n workflows as MCP tools (for Claude/Cursor to call). MCP Client node CALLS external MCP servers (like Hooklayer). You want the Client node.
How do I pass output from one Hooklayer tool to the next?
Use n8n expression syntax: {{ $json.recommended_chain[0].params.handle }}. analyze_account returns recommended_chain[] with pre-filled params; you can wire each step's params directly from the previous node's output.
Will n8n retry on transient failures?
Yes — n8n's built-in retry logic catches 5xx errors. Hooklayer returns 502/503 cleanly when upstream services fail (no credits charged). Configure retry-on-fail with a 3-attempt limit and exponential backoff for safety.
Can I rate-limit my n8n flows to stay within Hooklayer's tier?
Use n8n's "Wait" node between Hooklayer calls if you're on the Starter tier (60 req/min). Pro/Agency tiers (300 / 1,000 req/min) almost never need throttling — n8n's default rate is well below those limits.
How do I monitor Hooklayer credit burn in n8n?
Every Hooklayer response includes credits_remaining.total in the JSON. Add a small "IF" node after each MCP Client call: if credits_remaining.total < 100, send a Slack/email alert to top up.
Try it in n8n.
100 free credits at signup. No card. n8n setup in 60 seconds.
