llms.txt — AI-Readable API Reference
TCG API ships two plain-text files for AI agents and large language models, following the llmstxt.org convention:
/llms.txt— Spec-compliant index. H1 + summary + H2 sections of links pointing at our docs. Small enough to fit in any context window. Use this when you want an agent to discover what’s available./llms-full.txt— Single-file full reference. Every endpoint with parameters, tier requirement, x402 price, and response shape. Use this when you want an agent to build against the API without follow-up fetches.
Download llms.txt Short index — spec-compliant per llmstxt.org
Download llms-full.txt Full API reference for LLM ingestion
What’s in llms.txt
Section titled “What’s in llms.txt”A spec-compliant llms.txt is intentionally short. Ours has:
- H1 —
# TCG API - Blockquote — one-paragraph summary (games covered, tier prices, x402 networks)
- Paragraph — base URL and pointer to
llms-full.txt ## Documentation— links to introduction, quickstart, authentication, API explorer, llms-full.txt## API Reference— links to per-resource doc pages and the OpenAPI spec## Guides— errors, pagination, rate limits, x402## Pricing & Plans— pricing page + live x402 price map## Agent discovery— every.well-known/*profile (x402, ACP, UCP, MCP, RFC 9727 catalog)## Optional— pricing comparisons, per-game pages, contact
The ## Optional section has special meaning per the spec — its contents may be skipped when shorter context is needed.
What’s in llms-full.txt
Section titled “What’s in llms-full.txt”llms-full.txt inlines:
- Authentication for both X-API-Key and x402 (with USDC asset addresses on Base + Solana)
- Tier matrix — daily quotas, history range, bulk-resolve credit costs
- Response envelope and error format
- Every endpoint — path, method, parameters, tier, response, x402 price
- Data model — full field-level schemas for
Game,Set,Card,Price,PriceHistory,CardmarketPrice, bulk-resolve responses - x402 pay-per-request price table (atomic USDC + USD)
- Game slugs (T3 daily-update + T2 every-3-days)
- Agent discovery URLs
How to point your agent at this
Section titled “How to point your agent at this”Most LLM frameworks accept a URL or a chunk of context. Either:
Use the API documented at https://tcgapi.dev/llms-full.txt…or fetch the file once and paste it into your system prompt. It’s hand-maintained and tracks changes to openapi/spec.yaml, the worker’s PRICE_RULES, and the tier configuration in worker/src/types.ts.
For browser-based agents that speak WebMCP / SEP-1649, TCG API also exposes its tools via navigator.modelContext on every page — see the MCP server card for transport info.