Analook MCP Server

Community

Understand any competitor in 30 seconds — competitor intelligence for AI agents

How to install ↓ Compare GitHub source repository ↗

Category: Other #106 of 230
Stars: ★ 111 Popular
Transport: Streamable HTTP
Runtime: Hosted (no local runtime)
Credentials: API key / credential required
Last commit: 9 days ago
Tools: 5

55FMRS · C

Analook is the MCP surface of analook.com, exposing 5 tools over remote streamable-http for MCP clients and producing competitor teardowns from public data sources plus LLMs. It suits indie hackers, growth teams, and investors who need fast competitor intelligence; it is not a fit for teams requiring offline self-hosting or sensitive about data leaving to third-party services. An access token is required, and report conclusions should be reviewed by a human.

Strongest · Setup experience 13/20 Weakest · Maintenance 8/20

Reliability

9/20

Security and permissions

12/20

Maintenance

8/20

Documentation

13/20

Setup experience

13/20

Why each score

Reliability 9/20: Static evidence consists of a server.json manifest (streamable-http remote endpoint, version 1.0.2, carrying only schema/name/title/description/websiteUrl/repository/version/remotes with no declared capabilities, tools, or resources) plus a README listing five tools (analyze_competitor, get_report_status, get_report, get_report_markdown, list_my_reports). The repo description says FastAPI + Supabase, and the README references modules/mcp_app.py, tests/, and a Dockerfile, but no source files, tests, CI workflows, or mcp_app.py content were supplied for this review, so MCP init/tool-list handshake success, tool parameter/return fidelity, and error controllability cannot be verified. The manifest declares no tools while the README declares five, an information gap. Static calibration forbids reliability above 12 without committed tests and real CI; with no test, CI, or error-path evidence at all, 9 is warranted.
Security and permissions 12/20: The remote endpoint uses HTTPS (https://www.analook.com/mcp/) and the README shows a Bearer <YOUR_ANALOOK_TOKEN> placeholder rather than a real secret, so no 'install example with real credentials' red line is triggered; there is no direct evidence of malware, credential theft, or covert exfiltration. However, the server makes outbound calls carrying secrets to several third parties (DataForSEO Base64 email:password, Apify token, ProductHunt token, LLM keys), and the README describes credit gating, payment integrations, and Supabase auth/credits/promo codes — stateful operations that mutate quotas and can trigger billing — with no visible confirmation flow, least-privilege scoping, or data-flow disclosure. analyze_competitor ingests and stores third-party site data in Supabase with no stated retention or boundary policy. With key permission/confirmation/data-flow evidence missing, missing evidence is deducted rather than assumed optimal: 12, below the red-line band.
Maintenance 8/20: The repository is not archived, has 110 stars and 2 open issues, and server.json version 1.0.2 indicates some release iteration. But this material contains no commit history, release timeline, dependency-update record, issue-response evidence, or security-response channel. Licensing is visibly contradictory: repository metadata says License: unknown while the README embeds an MIT badge and states 'MIT' at the end — a governance defect in itself. The README references a pre-deploy preflight (scripts/verify_deploy_source.py) and a fly.toml app declaration, suggesting some operational discipline, but there are no governance files and no stated update path. Score 8.
Documentation 13/20: Coverage is reasonably broad: bilingual (EN/ZH) README, a module table mapping analysis modules to files and data sources, the seven-section AI insight structure, Quick Start with an explicit .env variable list, local run commands, Railway and Fly deploy steps, an API integration table, an external link to docs/mcp.html for token instructions, and manifest fields for websiteUrl and repository. Deductions: MCP tool parameters, return shapes, rate limits, and the concrete cost/quota rules of credit gating are barely documented in the visible content and require following an external page; self-hosting prerequisites such as applying Supabase migrations and configuring payments are not spelled out; there is no troubleshooting section. This is 'usable with hidden assumptions and troubleshooting gaps', so 13.
Setup experience 13/20: The remote path is clear: a standard mcpServers JSON snippet (url plus Authorization header) with a stated no-install/no-local-process model and five named tools, applicable to Claude Desktop, Cursor, and other mainstream clients. Self-hosting is also stepwise (git clone, pip install, .env, uvicorn app:app --port 8000) with a Railway one-click deploy and Dockerfile/Nixpacks notes. Missing: how to initialize Supabase (applying migrations), guidance for obtaining the five external API credentials, a fuller platform compatibility matrix beyond 'Python 3.13', and any committed evidence that a client connection was verified. Under static calibration setup may not exceed 15 without execution evidence, so 13.

Static review · not runListed 2026-09-10

Read the FMRS scoring method →

Fit and risk

What it can accessUses the network

Best for

Indie hackers who want competitor intelligence callable directly by an agent
Growth teams benchmarking against competitors
Investors doing quick pre-DD intelligence gathering
Founders preparing a launch in a crowded market

Not for

Teams requiring fully offline or self-hosted deployment with no data sent to third-party services
Self-hosted setups that cannot supply third-party API keys (such as DataForSEO, Apify, Product Hunt, and LLM providers)
Users expecting non-public internal data (this tool works from public signals only)
Clients that only support local stdio MCP processes and not remote HTTP MCP

Required permissions

Requires a valid Analook access token (Bearer token) for remote MCP authentication
Analysis depends on third-party data-source API keys (DataForSEO, Apify, Product Hunt, TeamoRouter, DeepSeek)
Server-side use of Supabase for auth, credits, and report storage

Risks and side effects

Reports are generated by third-party LLM and data services and may contain inaccurate or outdated information; verify key conclusions manually
Results are inferred from public signals, and business-model or revenue estimates are ranges and hypotheses only
Requests are sent to the remote service https://www.analook.com/mcp/, so data-transfer and compliance requirements should be assessed
A leaked token may allow others to consume your account credits
The repository declares no license metadata (the README states MIT); confirm authorization before commercial use

Setup

Before you start

Runtime：Hosted (no local runtime)

Obtain an Analook access token (see https://www.analook.com/docs/mcp.html for token instructions).
Add the remote server to your MCP client config:

{
  "mcpServers": {
    "analook": {
      "url": "https://www.analook.com/mcp",
      "headers": { "Authorization": "Bearer <YOUR_ANALOOK_TOKEN>" }
    }
  }
}

Save the config and restart your client.
This is a remote server — no local install or process is required.

claude_desktop_config.json

{
  "mcpServers": {
    "analook": {
      "url": "https://www.analook.com/mcp",
      "headers": { "Authorization": "Bearer <YOUR_ANALOOK_TOKEN>" }
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json

{
  "servers": {
    "analook": {
      "url": "https://www.analook.com/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_ANALOOK_TOKEN>"
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal

claude mcp add --transport http analook https://www.analook.com/mcp --header 'Authorization: Bearer <YOUR_ANALOOK_TOKEN>'

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

After adding the remote server at https://www.analook.com/mcp with a Bearer token in Claude Desktop or Cursor, the five tools (analyze_competitor, get_report_status, get_report, get_report_markdown, list_my_reports) should appear in the client's tool list; asking "list my reports" confirms the connection works.

Troubleshooting

Connection fails: confirm the client uses streamable-http and the URL is https://www.analook.com/mcp/
Unauthorized responses: check that the Bearer token in the Authorization header is valid and unexpired
Report stays in progress: use get_report_status to check the job, then retry or start a new analysis
Empty report: confirm the job is complete before calling get_report or get_report_markdown
Self-hosted analysis fails: verify all third-party service keys are set in .env

Things to try

Once connected, you can ask your AI assistant things like:

Analyze linear.app as a competitor using analyze_competitor
Check the progress of my report with get_report_status
Export the analysis of notion.so as Markdown with get_report_markdown
Show all my previous competitor reports with list_my_reports

Tools 5

analyze_competitor writes

Submit a competitor URL to start an analysis and generate a structured competitor teardown report.

get_report_status read-only

Check the current processing status of an analysis job.

get_report read-only

Retrieve the completed analysis report as structured data.

get_report_markdown read-only

Retrieve the completed analysis report in Markdown format.

list_my_reports read-only

List previously generated reports for the current account.

Use cases

Validate a market and competitive landscape before building

Generate traffic, SEO, and social benchmarks against competitors for growth teams

Collect quick competitor intelligence during pre-DD for investors

Find differentiated angles for a launch in a crowded niche

Supported clients

Claude Desktop

Cursor

Listed from the project's documentation, not tested by this site.

Overview

Analook is a remote MCP server that powers the analook.com backend, bringing competitor intelligence to AI agents: SEO, traffic, social, Product Hunt, pricing, and AI insights. Built on FastAPI + Supabase, it runs a 7-module parallel analysis pipeline that turns any product URL into a structured competitor teardown report. Any MCP-compatible client can connect over streamable-http at https://www.analook.com/mcp/ with no local install.

Similar servers

Hevy MCP 77 · B

Manage your Hevy workout data from AI assistants via MCP.

★ 517 · Tools 26 Compare with this →

Firecrawl MCP Server 75 · B

Firecrawl's official MCP server for web search, scraping, and structured extraction for AI agents

★ 7.6k · Tools 11 Compare with this →

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →

Compartment 75 · B

Encrypted, fully offline long-term memory for AI agents

★ 583 · Tools 14 Compare with this →

Compare with other MCP servers

Source revision 8740919ad589 Data synced 2026-10-19 Read the FMRS scoring method