LogNorm

npx add-mcp 'https://lognorm.com/api/mcp'

LogNorm MCP server

Hand your growth backlog to agents. LogNorm finds what's holding a website's growth back, ranks it, and lets AI agents such as Claude Code, Codex and Cursor do the work next to your team: fixes in your repo, posts with images, research. LogNorm re-checks every fix on the live site and measures what moved.

This repo holds the connection details, the LogNorm agent skill and a Claude Code plugin manifest for the hosted LogNorm MCP server.

Endpoint	https://lognorm.com/api/mcp
Transport	Streamable HTTP (remote)
Auth	OAuth 2.1 with dynamic client registration and PKCE. No API keys in prompts or config
Pricing	Free plan available. Agents run on your own Claude or Codex plan
Docs	https://lognorm.com/docs/agents

What agents can do with it

• Fix: read the site audit and GEO/AI-readiness audit (broken pages, metadata, indexing, AI-crawler access, llms.txt, structured data), fix the cause in your codebase, open a PR, then validate_fix so LogNorm re-checks the affected pages on the live site.
• Write: research briefs from the pages that rank, drafts in your voice, generated covers and diagrams, editor feedback loops.
• Plan: topics and hub-and-spoke clusters, and a ranked weekly plan of "moves" shared by people and agents.
• Research: keyword lookup and ideas, search volume and difficulty, competitors and the pages they win with, Google Search Console data.
• Get cited: track how ChatGPT, Gemini and Google AI Overviews answer your buyers' prompts and turn the gaps into moves.
• Leave a trail: every claim, comment, draft and fix shows on the LogNorm dashboard under the agent's name.

Tools include search, execute, claim_move, create_move, plan_move, update_move, validate_fix, start_run, research_brief, save_brief, save_draft, generate_image, plan_topic, keyword_lookup, add_competitor, add_ai_prompts, brain_add and more. Workflows are exposed as MCP prompts: weekly_growth, growth_review, fix_audit, write_post, plan_topic, ai_visibility.

Connect

The quickest route: paste this into your agent and it sets itself up.

Connect to LogNorm: follow https://lognorm.com/connect.md

Claude Code

claude mcp add --transport http --scope user lognorm https://lognorm.com/api/mcp

Then run /mcp, pick lognorm, Authenticate, and click Allow in the browser. Or install this repo as a plugin (MCP server + skill):

/plugin marketplace add lognorm/lognorm-mcp
/plugin install lognorm@lognorm

OpenAI Codex

codex mcp add lognorm --url https://lognorm.com/api/mcp
codex mcp login lognorm

Cursor

~/.cursor/mcp.json:

{ "mcpServers": { "lognorm": { "url": "https://lognorm.com/api/mcp" } } }

Claude (desktop or claude.ai)

Settings → Connectors → Add custom connector → https://lognorm.com/api/mcp → Connect → Allow.

Any other MCP client

Add a remote Streamable HTTP server named lognorm with URL https://lognorm.com/api/mcp. It answers 401 with OAuth metadata, so clients that support MCP authorization start the sign-in themselves.

Agent skill

skills/lognorm holds LogNorm's playbooks (fix audit issues, write posts, plan topics, run growth reviews). It contains instructions only, no credentials. Copy it into ~/.claude/skills, ~/.agents/skills or ~/.cursor/skills, or download https://lognorm.com/api/agents/skill.zip. Skill index: https://lognorm.com/.well-known/agent-skills/index.json

People without agents

Everything agents can do is also in the LogNorm web app for people, and API keys are available for scripts. Start free at https://lognorm.com/sign-up.

License

The skill and docs in this repo are MIT licensed. The LogNorm service is governed by its Terms of Service.