Looking for builders shipping workplace AI agents

[lililiu979-oss]

Pre-submission Checklist

• This post follows the MCP community guidelines

What would you like to share?

We’re opening the AgentArk builder whitelist for teams building workplace AI agents.

AgentArk is a marketplace where businesses discover and hire AI agents for real workflows across sales support operations recruiting finance and more.

If you’re building an agent that solves a real work problem we’d love to talk.

What builders get

• early onboarding
• • direct feedback while we shape listings and onboarding
• • 0% commission for the first 100 accepted builders for 6 months
    Would especially love to hear from builders working with MCP and related agent stacks.

Builder whitelist
https//docs.google.com/forms/d/e/1FAIpQLSfbBO3OkhL8ncDql1pcBJqYHkQKLDbi9SM2aXa2Zlv-jWofbQ/viewformusp=dialog

Relevant Links

No response

[armorer-labs]

One thing I keep seeing with workplace agents is that the hard part is usually not the first demo workflow. It is everything around trust and operations once the agent can touch real systems.\n\nThe recurring pain points seem to be:\n- where untrusted text can enter the run (tickets, docs, webpages, email, tool results)\n- how to gate high-risk actions without making the agent unusable\n- how to preserve auditability when actions span multiple tools and identities\n- how to keep policies understandable by the humans who have to approve exceptions later\n\nSo if anyone here is shipping into real workplace environments, I would love to hear which layer is causing the most friction right now: auth, approvals, context trust, tool safety, or post-hoc auditing. That seems like where the ecosystem can be most useful to each other.