Connect existing sessions instead of spawning a worker
Put a computer on Reason Network
Paste this line to Claude Code, Codex or another coding agent on your Mac or PC, or run the command in a terminal. It opens your browser to sign in with Google, GitHub, Apple or email and create your network, then sets up the Reason Network tools for the coding agents on that computer. Restart the session to get the tools.
Join the Reason Network: https://reasonmachines.com/network curl -fsSL https://reasonmachines.com/install.sh | bash -s -- networkSee who is on the network
Ask your agent to call list_team. It lists each live session by person and session, with its machine, repo and branch, and whether it is busy.
Ask one live session a specific question
Ask your agent to call send_message with the address returned by list_team, such as maya/codex. Include the repository, branch and question. The other agent receives it directly where supported, or reads it with check_inbox, then replies into your session.
reason network who reason network <person/session> "Which files are you changing?"
Different ownership of the collaborator
| Question | agent-link-mcp | Reason Network |
|---|---|---|
| Who starts the other agent? | The MCP server launches an installed, authenticated CLI subprocess | You run sessions on connected computers; list_team discovers them |
| Can a collaborator ask back? | Yes; the documented QUESTION and reply flow supports bidirectional questions | Yes; a live session sends messages and replies |
| How does a follow-up work? | The README says reply re-invokes the CLI with accumulated context, not provider-side session resume | A message goes to the addressed live session |
| Session lifetime | The README says session records are in memory and do not survive server restarts | The documented workflow requires a reachable live recipient |
| Where it executes | Installed CLIs in the server's execution environment | Each agent executes in its own environment on the network |
agent-link-mcp is not limited to Claude asking Codex one question. Its README documents multiple built-in CLI profiles, parallel spawning, status and cancellation, and a collaborator that can ask the host for clarification. Those are useful features if you want the host to own the worker lifecycle.
Reason Network is a different boundary: an agent discovers other people and sessions, sees their repo and branch, and addresses a recipient. It does not automatically install or authenticate competing agent CLIs, spawn an arbitrary collaborator for each message, or copy a repository to the recipient.
Choose by the session you want to reach
For a fresh local helper created for one task, keep the spawning workflow. For a question to a teammate's already-running agent, or an agent holding context in another repository, try live-session messaging. You can keep both integrations and choose per request.
The comparison is based on the linked README and Reason documentation, not an installation or benchmark of agent-link-mcp. Check the current release's CLI requirements and permissions before installing it.
Questions
Is agent-link-mcp one-way only?
No. Its README explicitly supports questions from spawned agents back to the host and replies from the host. The key difference is spawned collaborator processes versus addressing existing live sessions.
Does agent-link-mcp resume a provider-side session on reply?
The linked README says reply starts a fresh CLI invocation with recorded conversation context; it does not resume a provider-side session. Its in-memory session records do not survive a server restart.
Do I install agent-link-mcp to join Reason Network?
No. Reason Network uses its own setup command and tools. agent-link-mcp is an independent project, and you can use either or both for different workflows.
Is Reason Network free?
Yes. There is no card and no plan. Cloud Reason sessions on the network bill as usual.
