Agentzia
io.github.stoyan-stoyanovPublished by GitHub account stoyan-stoyanovv1.0.0Updated Oct 27, 2026
Test how AI assistants recommend your product and improve your landing page with evidence.
Overview
Agentzia runs simulated-buyer studies with ChatGPT, Claude, Gemini and Grok to show why AI assistants recommend competitors and how to improve a landing page.
- What you can do
- Find and read existing research runs, including progress, recommendations and supporting evidence
- Check account credits and advisory capacity before starting a study
- Start research on a public product page or on supplied public-safe marketing copy
- Resume eligible incomplete runs without spending another channel credit
- Create a human payment link for buying more credits, which the person completes in their browser
- When to use it
- Useful when you want evidence on how AI assistants describe and recommend your product versus competitors, and want your coding agent to turn that research into concrete landing-page changes and retests.
- Requirements
- A remote Streamable HTTP endpoint with OAuth 2.1 sign-in approved in a browser; an Agentzia account with credits (accounts include complimentary credits, and each selected assistant channel costs one credit). Works with any OAuth-capable Streamable HTTP client, or via the provided plugin and optional workflow skill.
Installation
In SourceWeft
- Open Agentzia in the dashboard and add it to a workspace.
- Connect the required credentials. They are stored privately and never shown on public pages.
- 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.
Use with an agent
Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.
Read https://sourceweft.com/mcp/io.github.stoyan-stoyanov%2Fagentzia/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.Other MCP clients
Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.
{
"mcpServers": {
"agentzia": {
"type": "http",
"url": "https://www.agentzia.dev/mcp"
}
}
}README
Agentzia plugins
Agentzia simulates buyers using ChatGPT, Claude, Gemini, and Grok to investigate why AI assistants recommend competing products and what to improve on a public landing page.
This public distribution package connects your coding agent to Agentzia's hosted MCP and includes the agentzia-optimize workflow skill. The hosted application and research engine are operated separately by Morphologic AI Inc.
Endpoint: https://www.agentzia.dev/mcp
Transport: Streamable HTTP
Authentication: OAuth 2.1 with browser approval
Cost: One channel credit per selected assistant; four channels by default. Accounts include complimentary credits. See current pricing.
Install
Codex
Run in a terminal:
Complete the client's OAuth sign-in when prompted, then start a fresh task if the running task has not loaded the plugin. Installing a package does not by itself authenticate your Agentzia account. See SETUP.md if tools are unavailable.
Claude Code
Run these commands inside Claude Code, one at a time:
Complete the Agentzia connection's browser sign-in when prompted. Use the client's MCP controls to authenticate or reconnect the plugin-provided server.
Cursor
The repository includes both an Agent Plugins manifest and a Cursor plugin manifest. For immediate MCP-only setup, add this to your project's .cursor/mcp.json and complete authentication in Cursor's MCP settings:
Install the optional workflow skill separately using the skills command below. Public marketplace review is separate from direct installation.
Gemini CLI
The extension declares the remote MCP connection. Complete the client's MCP authentication and reload/start a session as required by your installed version.
Portable Agent Plugins
plugin.json and mcp.json implement the Agent Plugins 1.0 format for compatible clients, including clients that support skills plus remote MCP. Follow your client's plugin installation and OAuth flow.
Skill only
If Agentzia is already connected:
The skill teaches the analyze, improve, and retest workflow. It does not install or authenticate the MCP connection by itself. OpenClaw users can use the same skill with their client's supported remote-MCP connection path.
MCP only
Any OAuth-capable Streamable HTTP client can connect directly, without this package. For example:
Use either the plugin connection or a direct connection to avoid duplicate tool sets. Client setup and troubleshooting.
Try it
- “Read my existing Agentzia report and explain the strongest evidence behind its recommendations.”
- “Analyze this public landing page with Agentzia. Use ChatGPT and Claude, with a budget of two channel credits.”
- “Use this Agentzia report to improve my landing page. Show which evidence supports each change; don't retest yet.”
The workflow checks existing research and credits, runs the requested study, reads compact results and supporting evidence, and helps your coding agent make changes. Retests use the user's agreed credit budget. Purchases remain in the human's browser.
Tools supplied by the hosted service
The server's current tool descriptions and documentation are the API contract. The skill adds the broader workflow without copying the full tool schemas.
Data and limitations
Submit public marketing pages or authorized public-safe copy. Inputs and traces are stored and processed by external research providers. Do not send secrets, private customer data, authenticated pages, or confidential drafts. Read the privacy policy, data handling information, and terms.
Results are observations of the tested scenarios, not guaranteed rankings or representative population estimates. Citation and reference-page checks do not establish participant visits or indexing. Incomplete coverage and comparisons between studies require care.
Maintainers
The canonical workflow skill lives in this repository. Core tool behavior stays documented and enforced by the hosted server. Update package versions and connection examples when their contracts change.
Publisher materials are in docs/listing.md; server.json describes the remote service for the Official MCP Registry. There are no local MCP processes, install hooks, bundled credentials, or research-model keys in this package.
The package is MIT licensed. Hosted Agentzia access is governed by its service terms and pricing. Support: [email protected].
Source: README.md at commit 6498108
Tools
0Version history
1- v1.0.0LatestSep 20, 2026
Score
85% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 daysRequired25 pts
- Auth declaredRequired10 pts
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts