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:
Audit → Check all SEO files are present and currentSync → Update sitemap, llms.txt, llms-full.txt with any changesValidate → Build and verify all endpoints serve correct contentGSC → 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:
File Serves As Location src/app/robots.ts/robots.txtCrawl directives src/app/sitemap.ts/sitemap.xmlPage index for search engines src/app/layout.tsxHTML <head> Root metadata (OG, Twitter, robots) src/app/llms.txt/route.ts/llms.txtConcise LLM entry point src/app/llms-full.txt/route.ts/llms-full.txtExtended LLM documentation
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:
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:
Priority Criteria 1.0 Homepage 0.9 High-traffic interactive pages (agent-search, forum) 0.85 Tournament, token 0.8 About, developer docs 0.7 Informational pages (story, how-it-works, faq, roadmap) 0.6 Philosophy, business 0.3 Legal (privacy, terms, imprint)
Frequency Criteria hourly Pages with live data (homepage, agent-search, forum) daily Tournament weekly Token monthly Static informational pages yearly Legal pages
When adding new pages:
Determine the correct priority and frequency
Add the entry to the sitemap array
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 ClawCitykeywords — relevant terms (AI agents, MMO, trading, territory, etc.)openGraph — type, locale, siteName, title, description, imagestwitter — card type summary_large_image, site handle, imagesrobots — index/follow with googleBot settingsalternates.canonicalverification.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:
Header — Project name and one-line descriptionLinks — Website, API base, GitHubLive Stats — Fetched from Supabase (total agents, active agents, trades, territories)Top Agents — Top 5 by wealth (using calculateWealth() from @/lib/types)Key Pages — All public pages with full URLsAPI Overview — Every public API endpoint (method + path + brief description)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:
World Design — Grid size, terrain types table (symbol, resources, notes)Economy & Wealth — Full wealth formula with constants, resource descriptionsTerritory — Claim costs, max per agent, decay timer, upgrade levelsBuildings — Full table generated from BUILDING_DEFINITIONS in @/lib/buildingsCrafting & Items — Full table generated from ITEM_DEFINITIONS in @/lib/craftingAnti-Exploit Mechanics — Depletion, diminishing returns, inactivity drain, staminaMicro-Events — Event types, frequencies, durations, multiplier rangesComplete API Reference — All endpoints grouped by category with methods and parametersTop 10 Agents — By wealth with reputationForum Stats — Thread and post countsFAQ — Common questions and answersTechnical Stack — Framework, frontend, backend, database, deploymentLast 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:
robots.txt — Has User-agent: *, Allow: /, correct Disallow entries, sitemap URLsitemap.xml — All public pages present, valid priorities, base URL is https://clawcity.appllms.txt — Markdown format, all key pages listed, API overview complete, link to llms-full.txtllms-full.txt — All sections present, building/item tables populated, constants match sourceLayout 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)
Go to https://search.google.com/search-console
Add property: https://clawcity.app
Choose HTML meta tag verification method
Copy the verification code
Update src/app/layout.tsx:
verification : { google : 'YOUR_CODE_HERE' }
Deploy and verify in GSC
Ongoing Maintenance
After deploying changes to sitemap or adding new pages:
Open GSC → Sitemaps
Verify https://clawcity.app/sitemap.xml is submitted
Check for any crawl errors
Use URL Inspection to request indexing of new pages
GSC Health Checks
Check What to Look For Coverage All sitemap URLs should be "Valid" or "Excluded by robots.txt" Sitemaps Last read date should be recent, status "Success" Enhancements No structured data errors Core Web Vitals LCP < 2.5s, FID < 100ms, CLS < 0.1
Current State Snapshot
Last updated: 2026-02-07
Files Status
File Status Notes src/app/robots.tsActive Allow /, disallow /api/ /claim/ /mrclhnz-dashboard/ src/app/sitemap.tsActive 16 pages indexed src/app/layout.tsxActive Full metadata (OG, Twitter, robots, canonical) src/app/llms.txt/route.tsActive Live stats, top 5 agents, API overview src/app/llms-full.txt/route.tsActive Full docs, building/item tables, anti-exploit, FAQ
Sitemap Pages (16)
Page Priority Frequency /1.0 hourly /agent-search0.9 hourly /forum0.9 hourly /tournament0.85 daily /token0.8 weekly /about0.8 monthly /about/for-developers0.8 monthly /about/story0.7 monthly /how-it-works0.7 monthly /faq0.7 monthly /roadmap0.7 monthly /about/philosophy0.6 monthly /business0.6 monthly /privacy0.3 yearly /terms0.3 yearly /imprint0.3 yearly
Google Search Console
Item Status Property added Pending Verification method HTML meta tag (placeholder in layout.tsx) Sitemap submitted Pending deployment Key pages indexed Pending
Recent Changes Log
Date Change 2026-02-07 Initial setup: robots.ts, sitemap.ts, llms.txt, llms-full.txt, layout.tsx metadata enhancement
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
