Nilyo
Sign inGet started →
Nilyo for persistent agents

Give OpenClaw structured access to your communication accounts

A persistent agent is a different risk profile from a chat: it runs without you watching, it reacts to events, and a mistake repeats. Nilyo is built for that case — every write is explicit, ambiguity blocks rather than guesses, and pacing is enforced on the server rather than trusted to the agent.

Add the server

Add Nilyo as an MCP server with the URL https://nilyo.com/mcp and OAuth, and install the Nilyo skill from the optional skills catalog for chaining guidance across providers. The first connection opens a browser once so you authorize the accounts the agent may use; after that it runs unattended.

Why an always-on agent needs a server-side budget

An agent that loops has no natural stopping point, and the account it drives is a real person’s. LinkedIn is capped at roughly a hundred actions per day per account, reads included, with calls serialized and spaced; WhatsApp caps brand new conversations at twenty a day. These are enforced where the calls are made, so an agent cannot decide to ignore them.

Events instead of polling

A persistent agent that polls every minute burns its budget on nothing. Realtime destinations deliver new messages and new emails to the agent when they happen, so it wakes on a real event with the identifiers it needs to fetch the conversation.

Writes stay deliberate

Reading never sends. A send is a distinct action, an ambiguous recipient blocks it, and a failed write is not retried — the state is read back. For an unattended agent that rule matters more than any capability: the failure mode of a retry loop is a contact receiving the same message five times.

“Notify me on WhatsApp when a LinkedIn message mentions pricing.”
“Every morning, summarize what came in overnight across my accounts.”

The same accounts elsewhere

The connection belongs to your Nilyo account. What OpenClaw uses unattended, you can inspect from ChatGPT or Claude the same day, with the same identifiers — useful when you want to check what the agent has been doing.

Common questions

What makes this suitable for a persistent agent?

Pacing and refusals live on the server, not in the prompt: budgets per account, serialized calls, blocked ambiguous writes and no blind retries. An agent cannot talk its way past them.

Does it poll or react?

It can react. Realtime destinations deliver new messages and emails as they happen, which is both cheaper and faster than polling.

Can it act on several accounts?

Yes, each action names the account it runs on, and one agent can hold LinkedIn, WhatsApp, several mailboxes and a calendar at once.

How do I stop it?

Disconnect the account or revoke the agent’s access from your account page; both cut it immediately, without touching your provider passwords.

How LinkedIn works with Nilyo

Connection. Secure hosted sign-in (your LinkedIn credentials go to LinkedIn, never to the agent). Sales Navigator and Recruiter are detected automatically when your account has them.

What your agent can do. People and company search, profile lookup from a name or URL, your network and invitations, private conversations (classic, Sales Navigator, Recruiter inboxes), posts, comments and reactions, job postings and applicants, Recruiter projects.

Pacing and safety. About 100 actions per day per account, profile views and searches included; calls are serialized so an agent can never run LinkedIn requests in parallel.

  • Profile URLs are resolved to stable profile IDs before any action.
  • Invitations carry an optional note of up to 300 characters.
  • Private or restricted profiles return an explicit “not accessible” result instead of a retry loop.

All LinkedIn guides

Related guides

Give Nilyo to your agent.

Paste one setup link. Your agent can guide OAuth, account connection and the first useful action.

© 2026 Nilyo.