# Hooklayer

> Hooklayer is the viral-content intelligence MCP for Claude power users. It exposes 100K+ analyzed viral TikTok, YouTube, and Instagram videos as 12 agent-callable tools organized into 5 layers (Score, Research, Analyze, Generate, Watch). Lives inside Claude Desktop, Cursor, n8n, and any MCP client, not as another SaaS dashboard tab. Test hooks, predict virality, find templates, analyze creators, remix viral URLs into fresh scripts, and track competitor creators for change detection over time.

Website: https://hooklayer.dev
Category: AI Developer Tool / MCP Server / Viral Content Intelligence
API Base: https://hooklayer.dev/api/v1
MCP Server: https://hooklayer.dev/api/mcp

## What Hooklayer Does

Hooklayer is the only MCP server in the Anthropic + Smithery directories positioned around viral-content TESTING, score before posting, predict before filming, find templates before scripting. Most creator AI tools generate content; Hooklayer tests it. The 12 tools fall into 5 workflow layers:

- **Score** (test before you post): score_hook, predict_virality
- **Research** (find what's working): trend_pulse, find_viral_template, search_videos
- **Analyze** (read creators + voices): analyze_account, match_voice
- **Generate** (produce in-voice scripts + full blueprints): viral_remix, brief_to_blueprint
- **Watch** (track competitors + detect changes over time): watch_account, list_watches, get_changes

Agents call Hooklayer tools to analyze TikTok, YouTube, and Instagram creators, score hooks against proven patterns, remix viral videos into fresh scripts, discover trending formats, match voice styles, predict virality, and monitor tracked creators for meaningful strategy shifts — all via REST API or MCP server.

## Safety and behavior

Most Hooklayer tools are research and analysis tools that may debit Hooklayer credits and record service usage. Only `list_watches` is fully read-only (pure DB query, no credit debit). `watch_account`, `list_watches`, and `get_changes` add non-destructive persistence for saved creator watches and historical snapshots. No Hooklayer MCP tool deletes user data or edits or deletes data on external social platforms (TikTok, YouTube, Instagram). `analyze_account` can return a `recommended_chain` field — an advisory list of possible follow-up tools with example parameters and reasons; the user or assistant decides whether to invoke them.

## 12 Tools (REST + MCP)

### analyze_account (5 credits)
- TikTok handle → full Viral DNA scores, format fingerprint, steal map
- Returns recommended_chain: pre-filled next tool calls for agentic loops
- 1-hour cache per handle

### score_hook (1 credit)
- Score a hook 0–100 against viral patterns
- Returns 3 rewrites, matched pattern, percentile ranking

### viral_remix (3 credits)
- URL or transcript → fresh script mirroring original viral DNA
- Preserves scene structure, energy pattern, camera directions

### trend_pulse (1 credit)
- Real-time rising opportunities + saturated patterns by niche
- Format/hook/style trends, not news topics
- 12-hour cache per niche

### find_viral_template (1 credit)
- Niche-fit ranked templates with hook patterns and format structures
- Returns example URLs and adaptation guidance

### match_voice (2 credits)
- Extract creator voice from reference samples
- Rewrite any draft in their style, returns voice signature

### predict_virality (2 credits)
- Score draft script for viral potential
- Returns viral DNA, retention diagnosis, target emotion, share reason

### search_videos (2 credits)
- Keyword search across TikTok, YouTube, and Instagram
- Ranked by weighted engagement (saves 8x, shares 5x, views 1x, comments 2x, likes 0.5x)
- Returns URLs, hook patterns, and creator handles for follow-up analysis

### brief_to_blueprint (7 credits)
- Turn a client brief or product description into a full ready-to-ship blueprint
- Chains templates + trends + hook_score + predict internally in one call
- Returns hook, script skeleton, format, and pre-scored virality prediction

### watch_account (0 or 5 credits)
- Start tracking a creator so future calls can detect what changed
- Creates a durable CreatorWatch + baseline CreatorSnapshot
- Costs 0 credits when a recent analyze_account result can be reused; otherwise 5
- Idempotent: repeat "track @same" calls dedupe on (user, platform, handle)

### list_watches (0 credits)
- List every creator the authenticated user is tracking
- Read-only DB query; never charges credits; never modifies state
- Returns compact watch metadata (id, platform, handle, snapshot_count, last_snapshot_at, is_active)

### get_changes (5 credits)
- Detect what meaningfully changed since the previous saved baseline for a tracked creator
- Runs a fresh analyze under the hood; appends a new CreatorSnapshot; snapshotCount +1
- Returns only high/medium importance shifts (viral DNA, engagement, format, hook pattern, topic, outliers)
- Returns status=no_meaningful_change honestly when nothing shifted

## Authentication

Two auth modes:

1. **Bearer hl_live_\* keys**: minted in the dashboard at https://hooklayer.dev/dashboard. Free tier signup grants a key + 25 lifetime credits, no card required.
2. **OAuth 2.1 + PKCE**: full discovery + Dynamic Client Registration. Endpoints below.

## OAuth 2.1

- Authorization Server Metadata (RFC 8414): https://hooklayer.dev/.well-known/oauth-authorization-server
- Protected Resource Metadata (RFC 9728): https://hooklayer.dev/.well-known/oauth-protected-resource
- Dynamic Client Registration (RFC 7591): POST https://hooklayer.dev/oauth/register
- PKCE: S256 challenge method
- 401 responses include WWW-Authenticate header pointing at resource metadata for automatic client discovery

## Pricing

- Free: $0/month, 25 lifetime credits, 30 req/min (no card required)
- Hobby: $19/month or $180/year, 1,200 credits/month, 45 req/min, email support
- Starter: $49/month or $468/year, 3,200 credits/month, 60 req/min, email support
- Pro: $149/month or $1,428/year, 10,000 credits/month, 300 req/min, brand voice memory, priority support [Most popular]
- Agency: $499/month or $4,788/year, 32,000 credits/month, 1,000 req/min, white-label, 10 seats, custom hook fine-tune, dedicated Slack support
- Pay-as-you-go: $10 for 400 credits or $25 for 1,200 credits, never expire
- No seat charges, no AI Maker tax

## Integration

- Install page (all clients): https://hooklayer.dev/mcp
- REST API base: https://hooklayer.dev/api/v1
- MCP server endpoint: https://hooklayer.dev/api/mcp (drop into Claude Desktop, Cursor, n8n in 30 seconds)
- Playground: https://hooklayer.dev/playground (5 free calls per hour per IP, no signup)
- GitHub repo (public examples): https://github.com/khan-ashifur/hooklayer

## Discovery

- Listed on the Anthropic Official MCP Registry: `io.github.khan-ashifur/hooklayer` v1.0.0
- Search: https://registry.modelcontextprotocol.io/v0/servers?search=hooklayer
- Auto-propagates to PulseMCP and other registry aggregators

## Who Uses Hooklayer

- AI agent builders adding content intelligence to workflows
- n8n / Make / Zapier automations needing a viral-content quality gate
- Content teams building agentic loops for TikTok and short-form video
- Developers integrating hook scoring and trend data into apps

## Links

- Homepage: https://hooklayer.dev
- Documentation: https://hooklayer.dev/docs
- Best Viral Content Research Tools 2026 (flagship comparison + listicle): https://hooklayer.dev/best-viral-content-research-tools-2026
- All 12 MCP tools (see /tools for the reference pages that have long-form docs; the retention tools — watch_account, list_watches, get_changes — currently ship without dedicated marketing pages): https://hooklayer.dev/tools
- All client integrations: https://hooklayer.dev/for
- Build guides (outcome-oriented tutorials): https://hooklayer.dev/build
- Pricing: https://hooklayer.dev/pricing
- Playground: https://hooklayer.dev/playground
- FAQ: https://hooklayer.dev/faq
- Blog: https://hooklayer.dev/blog
- vs Composio TikTok MCP: https://hooklayer.dev/vs/composio-tiktok
- vs Apify TikTok Scraper: https://hooklayer.dev/vs/apify-tiktok
- vs Generic TikTok MCPs (seym0n/tiktok-mcp etc.): https://hooklayer.dev/vs/tiktok-mcp
- Get API Key: https://hooklayer.dev/auth/signup
- Privacy Policy: https://hooklayer.dev/privacy
- What's new in v1.1: https://hooklayer.dev/whats-new

## Tool Reference Pages

- analyze_account: https://hooklayer.dev/tools/analyze-account
- score_hook: https://hooklayer.dev/tools/score-hook
- viral_remix: https://hooklayer.dev/tools/viral-remix
- trend_pulse: https://hooklayer.dev/tools/trend-pulse
- find_viral_template: https://hooklayer.dev/tools/find-viral-template
- match_voice: https://hooklayer.dev/tools/match-voice
- predict_virality: https://hooklayer.dev/tools/predict-virality
- search_videos: https://hooklayer.dev/tools/search-videos
- brief_to_blueprint: https://hooklayer.dev/tools/brief-to-blueprint (integrated on the /brief-to-blueprint page)
- watch_account, list_watches, get_changes: retention MVP tools; discoverable via MCP tools/list, docs summarized here and at /docs (dedicated reference pages ship in a follow-up)

## Build Guides (outcome-oriented, chain 2-3 tools)

- Build a viral TikTok research pipeline: https://hooklayer.dev/build/viral-tiktok-research-pipeline
- Build an AI content quality gate: https://hooklayer.dev/build/ai-content-quality-gate
- Build a TikTok competitor analysis bot: https://hooklayer.dev/build/competitor-analysis-bot

## Client Integration Pages

- Hooklayer for Claude Desktop: https://hooklayer.dev/for/claude-desktop
- Hooklayer for Cursor: https://hooklayer.dev/for/cursor
- Hooklayer for n8n: https://hooklayer.dev/for/n8n
- Hooklayer for OpenAI Agents SDK: https://hooklayer.dev/for/openai-agents-sdk
- Hooklayer for LangChain: https://hooklayer.dev/for/langchain
- Hooklayer for Windsurf: https://hooklayer.dev/for/windsurf

## Long-Form Reading

- MCP vs REST for content agents: https://hooklayer.dev/blog/mcp-vs-rest-for-content-agents
- How to analyze any TikTok account: https://hooklayer.dev/blog/analyze-tiktok-account-with-claude
- The Viral DNA Method (6 scores): https://hooklayer.dev/blog/viral-dna-method
- n8n + TikTok MCP: the 3-step viral pipeline: https://hooklayer.dev/blog/n8n-tiktok-mcp-viral-pipeline
- How to add Hooklayer to Claude Desktop (claude_desktop_config.json template): https://hooklayer.dev/blog/add-hooklayer-to-claude-desktop
- Best MCP servers for AI content workflows in 2026: https://hooklayer.dev/blog/best-mcp-servers-2026