Run locally
Overloop CLI
Use a terminal or a local coding agent to read an existing campaign and its metrics. The quickstart uses two read-only calls.
Open the tested quickstart →Overloop / Developers
Read campaign performance from Overloop with a local CLI. Give Codex or Claude Code a bounded task, inspect the JSON result and decide what to improve.
Run locally
Use a terminal or a local coding agent to read an existing campaign and its metrics. The quickstart uses two read-only calls.
Open the tested quickstart →Build an integration
The CLI uses the Overloop API v2. Check authentication, account access and the operation you need before implementing a direct integration.
Read the API reference →Inspect signal types
Max, part of Overloop, publishes a local stdio MCP server. Our SDK test listed its tools and read the public signal catalog.
See the tested operation →The working Overloop example here uses a CLI. A native Overloop MCP connection is not part of this quickstart.
The campaign-read scenario was executed in Codex's local shell and independently through Claude Code. Both used the same pinned CLI and an existing authenticated account. No campaign was changed or launched.
The next step is concrete: retrieve an existing campaign, read its channel metrics and review the result. This checks the connection before you expand into prospect research or outreach preparation.
Use the Codex and Claude Code task →Separate company search, contact lookup and enrichment before choosing a provider.
Data connection guide →Review the event, source date and account fit before treating a signal as a reason to contact someone.
Buying signals playbook →Distinguish reading a campaign from creating, enrolling, activating or sending.
Outreach operation guide →The official Max MCP server is published as sortlist-max-mcp. On September 23, 2026, we tested package 1.0.2 with an MCP SDK stdio client and Node 25.6.1: initialization succeeded, tools/list returned 19 tools, and one list_signals call returned 18 public signal definitions.
This gives a builder a first check: inspect the available signal types before designing a monitoring workflow. The catalog lists definitions, not prospects. It does not prove account access, plan entitlement or lead quality.
Download the dated MCP test receipt (JSON) →
The SDK test used a non-secret placeholder for the public catalog. Native Codex and Claude setup remains unverified; a separate Codex CLI attempt stopped at its tool-approval policy. Authenticated lead retrieval, webhook delivery and outreach need separate verification.
Use the Max API reference to check account operations, or the B2B data MCP comparison to evaluate other data connectors.
Max / yourmax.ai is part of Overloop. Its CLI package is sortlist-max-cli, with the command max. Authentication and supported operations differ from the Overloop CLI. The campaign-reporting test above does not verify a Max signal-retrieval workflow.
Use the buying-signals overview to check the prospecting use case and the repository documentation to inspect the current interface.