1NFER Inference Discovery

business.hesoyamv1.0.0Updated Oct 10, 2026

Inspect 1NFER models, crypto funding terms and task cost without keys or paid calls.

VerifiedStreamable HTTPWeb executableAI & MLFinance

Overview

AI-generated overview

A public read-only MCP server for inspecting 1NFER model listings, crypto payment terms and task cost estimates without keys or paid calls.

What it does
Exposes four read-only tools over Streamable HTTP: list_models, payment_terms, estimate_cost and connection_recipe. It lets an assistant look up catalogue model aliases, current funding terms and cost estimates for a task, and retrieve a connection recipe. It performs no model generation, creates no invoices and takes no payments on the public endpoint.
When to use it
Useful when you want an assistant to check what 1NFER models exist, what the current crypto funding terms are, or what a task would cost before committing. Also handy for getting connection details for the service. It is not for running inference or handling payments.
Requirements
The public server is a remote Streamable HTTP endpoint; no API key is required or accepted and no environment variables or headers are declared. A separate local mode exists: install pinned Python dependencies in an isolated environment and run the server with the stdio flag. The local mode optionally uses INFER_STATE_DIR for a private state directory.
Before you install
The public endpoint is read-only and takes no payments. The local mode adds account and invoice tools; invoice creation needs an authorized amount and a saved intent ID, and it never signs or sends cryptocurrency. Keys and saved invoice intents are stored in the state directory, which is an API credential, not a crypto wallet seed. Read live payment terms and your account's actual rates before use, and reconcile unknown outcomes rather than starting a new transfer.

Installation

In SourceWeft

  1. Open 1NFER Inference Discovery in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "inference-discovery": {
      "type": "http",
      "url": "https://hesoyam.business/agent-tools/mcp"
    }
  }
}

README

1NFER agent connection tools

Public read-only MCP: https://hesoyam.business/agent-tools/mcp (Streamable HTTP). Tools: list_models, payment_terms, estimate_cost, connection_recipe. No API key required or accepted, no model generation, no invoice creation or payments on this public server.

Local procurement: install the pinned Python dependencies into an isolated environment and run python server.py --stdio. Local-only extra tools: register_account, account_status, create_invoice, invoice_status. INFER_STATE_DIR optionally selects your private local directory (default ~/.config/1nfer-client, mode0700). Keys and saved invoice intents are0600; secrets are not tool outputs. Preserve this directory for account access. This is an API credential, not a crypto wallet seed.

Invoice creation requires an authorized amount and saved intent ID. It never signs or sends cryptocurrency. Do not start a new transfer after expiry; wait for confirmed credit. Unknown outcomes require reconciliation, not a new key/payment. Read live payment terms and your account's actual rates before use.

Installable skills are under skills/. Existing inference: https://1nfer.net/ . Scoped own-brand business setup: https://hesoyam.business/router . No guaranteed customers, returns, free model or full-framework compatibility. All named models are exact catalogue aliases; provider identity alone does not establish provenance.

Source: README.md at commit 5937656

Tools

0
Tool metadata has not been indexed yet.

Version history

1
  1. v1.0.0LatestOct 10, 2026