Negotiator MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Negotiator MCP Serverfind me the best price on a PS5 and draft messages to sellers"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Negotiator MCP Server
Built for the Amazon Developer Hackathon "Build, Ship, Shape" — Alexa+ track (real cash: $25,000 / $15,000 / $4,000 for 1st/2nd/3rd, plus AWS credits; deadline Oct 23, 2026, 2:00pm CDT).
Why this exists
The Alexa+ track requires a self-hosted MCP server (spec 2025-11-25 or later, Streamable HTTP transport) or an Agent Skill — per the rules, "MCP is the open standard that powers Alexa+ integrations." (Model Context Protocol, Alexa+ docs)
This server exposes logic already proven twice elsewhere in ArkNet's real projects, now behind the exact protocol this hackathon requires:
AI-NEGOTIATOR (live at ainegotiator.digital) — turns a plain-language purchase request into ranked, cheapest-first market-price estimates with negotiation scripts, via Groq (
openai/gpt-oss-20b).negotiator-agent (built for the Agents for Humans hackathon) — added a second step: drafting a ready-to-send outreach message per vendor.
Same model, same prompts, same "AI-generated estimate, not live pricing" disclosure — just re-exposed as a standard MCP server so Alexa+ (or any MCP client) can call it directly: "find me the best price on X and draft messages to the sellers," handled end to end.
Related MCP server: agent-commerce-mcp-server
Tools
find_and_rank_offers(purchase_request, location, currency)— ranked, cheapest-first vendor estimates with negotiation scripts.draft_negotiation_message(vendor, negotiation_script, buyer_name)— turns a script into a ready-to-send message.
Architecture
flowchart LR
A[Alexa+ / any MCP client] -->|MCP, Streamable HTTP| S[Negotiator MCP Server]
S -->|tool call| R[find_and_rank_offers<br/>Groq openai/gpt-oss-20b]
S -->|tool call| D[draft_negotiation_message<br/>Groq openai/gpt-oss-20b]
R --> S
D --> S
S --> ARun it
pip install -r requirements.txt
export GROQ_API_KEY=... # same key AI-NEGOTIATOR's Edge Function uses
python server.py # serves MCP over Streamable HTTPStatus
Scaffolded by an automated weekly hackathon-scout routine on Joy Ogunleye's behalf. Not yet run against a live Groq key, and not yet registered as an actual Alexa+ Agent Skill in this session — no credentials were available to the automation that built this, and Alexa+ developer preview access needs to be requested by Joy directly. Not submitted to Devpost. See the tracking issue in AI-NEGOTIATOR for the full build plan and remaining steps before submission.
License
Apache 2.0 — see LICENSE.
This server cannot be deployed
Maintenance
Related MCP Connectors
AI-powered product search, affiliate links, and price negotiation for e-commerce platforms
Deep purchase intelligence for agents and apps: live prices, landed cost, trust scores. Free.
Autonomous commerce for AI agents: discover, quote, order, pay, verify.
Negotiation math engine. Pareto frontier, counteroffer generation, zero LLM tokens.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceIntelligently generates cost estimates and lead times for manufacturing RFPs by parsing requests, matching against historical quotes, and calculating activity-based costs with confidence scoring and human approval workflows.-
- AlicenseAqualityDmaintenanceEnables AI agents to create, compare, and track purchases with structured buying workflows, offer comparison, and merchant verification.5MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to dispatch legal vendor requests (court reporters, experts, e-discovery, etc.) and receive bids within minutes, with tools for matter management and negotiation.1Apache 2.0
- FlicenseAqualityDmaintenanceEnables AI-powered shopping assistance by analyzing natural language shopping queries and automating product searches on multiple e-commerce platforms.13-