AGENT DOCUMENTATION HUB

AGENT-READABLE
ENTRYPOINTS AND TOOLS

This page is the public docs hub for autonomous and coding agents evaluating The Aggregator. It points to the machine-readable files, confirmed MCP tools and resources, and the onboarding path that already exist.

SAFETY AND OPERATING BOUNDARIES

Use sandbox or test keys before any production credentials or go-live checks.

Treat app and admin actions as authenticated surfaces that may require explicit human approval.

Use public discovery assets for orientation, then move into the app flow for registration, API keys, and controlled rollout.

/llms.txt

Short machine-readable index for LLMs and agents starting discovery on the public host.

/llms-full.txt

Expanded machine-readable inventory with route summaries, package surfaces, and integration entrypoints.

@aggregator-gg/mcp-server

Confirmed MCP package for Claude Code, Cursor, and Codex-style tooling with catalog, session, transaction, and provider tools.

TypeScript SDK (coming soon)

TypeScript SDK — in development, not yet published to npm. Use the REST API or the MCP server today.

CONFIRMED MCP TOOLS

list_games

get_game

create_session

create_demo_session

get_session

list_transactions

list_providers

provider_health

CONFIRMED MCP RESOURCES

aggregator://docs/quickstart

aggregator://docs/callbacks

aggregator://docs/api-reference

These resources expose quickstart, callback, and API reference material directly through the MCP server so coding agents can stay inside one tool surface.

START FROM THE PUBLIC CONTRACT

If you are evaluating integrations autonomously, begin with llms.txt, inspect the MCP and SDK surfaces, and only then move into sandbox or authenticated setup.

/LLMS.TXT integration.downloadAgent For Agents

AGENT-READABLE ENTRYPOINTS AND TOOLS