seo-llms

name: seo-llms description: Audit and update SEO infrastructure (robots.txt, sitemap.xml, llms.txt, llms-full.txt) and manage Google Search Console. Run after adding pages, changing routes, or updating game mechanics.

SEO & LLMs Agent

Maintains SEO infrastructure and LLM discoverability for ClawCity. Run this agent after adding new pages, changing routes, updating game mechanics, or when preparing for search engine submissions.

Quick Start

When triggered, execute these phases in order:

1. Audit → Check all SEO files are present and current
2. Sync → Update sitemap, llms.txt, llms-full.txt with any changes
3. Validate → Build and verify all endpoints serve correct content
4. GSC → Check Google Search Console status (if configured)

Phase 1: Audit

Files to Check

Read these files and verify they exist and are up-to-date:

Cross-Reference: Discover All Public Pages

Scan for all public (non-API, non-admin) page routes:

src/app/**/page.tsx  (excluding api/, mrclhnz-dashboard/, claim/)

Compare discovered pages against the sitemap entries in src/app/sitemap.ts.

Flag any pages missing from the sitemap.

Cross-Reference: Discover All API Routes

Scan for all API route files:

src/app/api/**/route.ts

Compare against the API sections in both llms.txt and llms-full.txt.

Flag any endpoints missing from the LLM files.

Phase 2: Sync

2a. robots.ts

Verify rules match current private routes:

// Expected structure
disallow: ['/api/', '/claim/', '/mrclhnz-dashboard/']
sitemap: 'https://clawcity.app/sitemap.xml'

Check for new private routes that should be disallowed (admin panels, internal tools, etc.).

2b. sitemap.ts

For each public page, ensure an entry exists with appropriate priority and frequency:

When adding new pages:

1. Determine the correct priority and frequency
2. Add the entry to the sitemap array
3. Keep entries sorted by priority (descending)

2c. layout.tsx Metadata

Verify the root metadata object includes:

• metadataBase: new URL('https://clawcity.app')
• title with template format: { default: '...', template: '%s | ClawCity' }
• description — accurate summary of ClawCity
• keywords — relevant terms (AI agents, MMO, trading, territory, etc.)
• openGraph — type, locale, siteName, title, description, images
• twitter — card type summary_large_image, site handle, images
• robots — index/follow with googleBot settings
• alternates.canonical
• verification.google — filled in once GSC is set up (may be commented out)

If the project description, name, or branding changes, update all metadata fields to match.

2d. llms.txt

This is the concise entry point for LLM crawlers. It should contain:

1. Header — Project name and one-line description
2. Links — Website, API base, GitHub
3. Live Stats — Fetched from Supabase (total agents, active agents, trades, territories)
4. Top Agents — Top 5 by wealth (using calculateWealth() from @/lib/types)
5. Key Pages — All public pages with full URLs
6. API Overview — Every public API endpoint (method + path + brief description)
7. Link to llms-full.txt — For extended documentation

Data Sources

import { createServerClient, isSupabaseConfigured } from '@/lib/supabase';
import { calculateWealth } from '@/lib/types';

• Always guard with isSupabaseConfigured — gracefully degrade if DB unavailable
• Use revalidate = 3600 for 1-hour ISR cache
• Return Content-Type: text/plain; charset=utf-8

When API endpoints change, update the API Overview section.

2e. llms-full.txt

This is the extended guide. It should contain everything from llms.txt plus:

1. World Design — Grid size, terrain types table (symbol, resources, notes)
2. Economy & Wealth — Full wealth formula with constants, resource descriptions
3. Territory — Claim costs, max per agent, decay timer, upgrade levels
4. Buildings — Full table generated from BUILDING_DEFINITIONS in @/lib/buildings
5. Crafting & Items — Full table generated from ITEM_DEFINITIONS in @/lib/crafting
6. Anti-Exploit Mechanics — Depletion, diminishing returns, inactivity drain, stamina
7. Micro-Events — Event types, frequencies, durations, multiplier ranges
8. Complete API Reference — All endpoints grouped by category with methods and parameters
9. Top 10 Agents — By wealth with reputation
10. Forum Stats — Thread and post counts
11. FAQ — Common questions and answers
12. Technical Stack — Framework, frontend, backend, database, deployment
13. Last Updated — ISO timestamp

Data Sources

import { createServerClient, isSupabaseConfigured } from '@/lib/supabase';
import { calculateWealth, WORLD_SIZE, WEALTH_SCALE_FACTOR, ... } from '@/lib/types';
import { BUILDING_DEFINITIONS, BUILDING_DECAY_HOURS } from '@/lib/buildings';
import { ITEM_DEFINITIONS } from '@/lib/crafting';

When game mechanics change (new buildings, items, terrain types, constants), the tables in llms-full.txt auto-update because they're generated from source code. But verify:

• New API endpoints are added to the Complete API Reference
• New game systems get their own section
• FAQ is updated for new features
• Constants imported from types.ts are still accurate

Phase 3: Validate

Build Check

npm run build

Verify no TypeScript or build errors. Check the build output for:

• /robots.txt — listed as static route
• /sitemap.xml — listed as static route
• /llms.txt — listed with ISR revalidation
• /llms-full.txt — listed with ISR revalidation

Also verify workspace conventions remain valid:

npm run check:agent-layout

Content Verification

After a successful build, verify content correctness by reading the source:

1. robots.txt — Has User-agent: *, Allow: /, correct Disallow entries, sitemap URL
2. sitemap.xml — All public pages present, valid priorities, base URL is https://clawcity.app
3. llms.txt — Markdown format, all key pages listed, API overview complete, link to llms-full.txt
4. llms-full.txt — All sections present, building/item tables populated, constants match source
5. Layout metadata — OpenGraph and Twitter tags present in page source

Local Testing (if dev server available)

curl -s http://localhost:3000/robots.txt | head -10
curl -s http://localhost:3000/sitemap.xml | head -20
curl -s http://localhost:3000/llms.txt | head -30
curl -s http://localhost:3000/llms-full.txt | wc -l

Phase 4: Google Search Console

Initial Setup (one-time)

1. Go to https://search.google.com/search-console
2. Add property: https://clawcity.app
3. Choose HTML meta tag verification method
4. Copy the verification code
5. Update src/app/layout.tsx:

   verification: { google: 'YOUR_CODE_HERE' }
   

6. Deploy and verify in GSC

Ongoing Maintenance

After deploying changes to sitemap or adding new pages:

1. Open GSC → Sitemaps
2. Verify https://clawcity.app/sitemap.xml is submitted
3. Check for any crawl errors
4. Use URL Inspection to request indexing of new pages

GSC Health Checks

Current State Snapshot

Last updated: 2026-02-07

Files Status

Sitemap Pages (16)

Google Search Console

Recent Changes Log

Agent Execution Summary

┌─────────────────────────────────────────────────────────┐
│  SEO & LLMs AGENT                                       │
├─────────────────────────────────────────────────────────┤
│                                                         │
│  1. AUDIT:                                              │
│     • Read all 5 SEO files                              │
│     • Scan src/app/**/page.tsx for public pages         │
│     • Scan src/app/api/**/route.ts for API endpoints    │
│     • Compare against sitemap + llms.txt + llms-full    │
│                                                         │
│  2. SYNC:                                               │
│     • Add missing pages to sitemap.ts                   │
│     • Add missing endpoints to llms.txt/llms-full.txt   │
│     • Update constants if game mechanics changed        │
│     • Update layout.tsx metadata if branding changed    │
│     • Update robots.ts if new private routes added      │
│                                                         │
│  3. VALIDATE:                                           │
│     • npm run build — no errors                         │
│     • robots.txt — valid directives                     │
│     • sitemap.xml — all public URLs present             │
│     • llms.txt — markdown with live stats               │
│     • llms-full.txt — complete game documentation       │
│                                                         │
│  4. GSC:                                                │
│     • Check verification status                         │
│     • Submit sitemap if not submitted                   │
│     • Request indexing of new pages                     │
│     • Report any crawl errors                           │
│                                                         │
│  5. UPDATE this snapshot with changes made              │
│                                                         │
└─────────────────────────────────────────────────────────┘

Notes for Agent

• Always guard Supabase calls with isSupabaseConfigured — the app should work without a DB
• Use ISR (revalidate = 3600) for llms routes — no need for real-time data
• Return text/plain content type for llms routes, not JSON
• Building and item tables are auto-generated from source code definitions
• When importing from @/lib/types, cast item definitions if needed to satisfy TypeScript strict mode
• The verification.google field in layout.tsx is commented out until GSC setup is complete
• Always update the "Current State Snapshot" section after making changes