Connect
Agent-legible
Built to be read by agents, not just people.
We build growth for the age of AI, so we built the site to be legible to it. Point an agent at Pollyester and it reads what you read: our services, how we work, what we publish, and the roles we're hiring for. Same content, in a form machines can use.
Connect an agent
One endpoint. Read-only.
We run a Model Context Protocol server. Add it as a connector in your AI tool and the agent can search and fetch the agency directly. It reads. It never writes, submits, or acts on your behalf.
- Our services, and what each one does
- The growth method, wave by wave
- The writing we publish
- The roles we're hiring for
# Streamable HTTP · read-only · no auth
https://www.pollyester.com/api/mcp
# or, in Claude Code
claude mcp add --transport http pollyester https://www.pollyester.com/api/mcp
In Claude
Open Settings, then Connectors. Add a custom connector and paste the endpoint. In Claude Code, run the command above.
In ChatGPT
Turn on connectors, add a remote MCP server, and paste the endpoint. The tools are read-only, so nothing on your side is touched.
For crawlers and tools
The same content, published plainly.
Not everything speaks MCP. Append .md to any page URL for its markdown twin, and start from these:
- /llms.txtthe site, mapped for language models
- /llms-full.txtthe same map, with the writing inlined
- /sitemap.xmlevery canonical URL
Let's talk.
Tell us where you're trying to grow. We'll tell you the truth about how to get there.