Outreach MCP with Claude: sequences and meetings from Outreach, competitor data from Linkeddit

It gives Claude Outreach records and conversations: prospects, accounts, opportunities, users, sequences, emails and Kaia meeting records with transcripts and summaries. The Agent Index lists 15 tools for the Claude connector, all classed as reads, with permissions marked read only.

Outreach's own documentation describes a wider set. Its help center lists tools to search Kaia meetings, search sequences and the state of prospects in them, search tasks, and ask questions about an account or opportunity, plus create and delete tools. Its developer portal says the catalog has 32 tools. The two sources differ, so check which tools your Claude connector shows before you plan a workflow around one.

One gate applies to all of it. The help center says a user needs an active, licensed seat and the Amplify add-on with active credits, and an Outreach admin has to switch the server on. Outreach states that it has not exposed any resources or prompts, only tools.

Outreach knows your prospects, your sequences and what was said in your meetings. It does not know what your competitors published, changed or were criticised for, or which brands the answer engines recommend when a buyer asks for a tool in your category. Linkeddit measures those three things and returns them as data, with no advice attached.

Compete tracks competitor moves with a priority and a reason on each, and stores competitor reviews from G2, Capterra, TrustRadius and Trustpilot. Answer Radar records where ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews name a competitor instead of you. Both come with one plan at $60 a month. Linkeddit does not find or contact prospects, and it sends nothing.

Sellers describe what goes wrong without it. A founder in a sales community wrote that "vague “we do it differently” claims kill cold outreach." A Claude user on Reddit put the same point about battlecards: "A battle card that ignores competitor strengths is useless in a real sales conversation." Specific, current, sourced facts are what fix both.

Ask Claude to read how your current sequences and recent emails describe the competitor, then ask Linkeddit what that competitor changed and what its customers complain about, and have Claude draft revised messaging with a source on every claim. You review the draft and edit the sequence yourself in Outreach.

A prompt that works: "Use the Outreach connector to find the sequences and recent emails that mention Competitor X. Then use the Linkeddit connector to get Competitor X's recent signals and its reviews rated two stars or lower. Draft a replacement for the second email in each sequence that makes one specific, accurate point, with the source link for each claim. Do not claim anything the data does not show, and do not change anything in Outreach."

Keep the last sentence of that prompt. Outreach's help center says create actions are on by default and delete actions are off by default, and its documented tools include creating sequences and adding prospects to them. Writing a draft in chat costs nothing, but a mistaken change to a live sequence reaches buyers. Telling Claude to draft only keeps a person between the data and a send.

On Linkeddit's own test project, plausible.io, a read-only run returned five competitor signals for the week of 2026-10-26, all low priority, from Simple Analytics and Umami. One was an Umami post previewing what is coming in its v3 release, with the reason it matters: visibility into its upcoming architectural or feature milestones.

The same project's answer gaps returned one open gap: the buying question "Umami vs Matomo for self-hosted analytics", where the project is not represented, with a confidence of 1 and an evidence score of 22. The reviews view returned three one-star Trustpilot reviews of plausible.io, one dated 2026-03-23 about an unresolved tracking issue and two about locked dashboards and billing.

That is what Claude would hand to a seller: a dated competitor move, a question where the engines name someone else, and what unhappy customers wrote. The test project is an analytics tool, not a sales team, so the point is the shape of the evidence, not these specific results. These tools ran as read-only calls. The Outreach side was not run.

Linkeddit shows which pages the engines cite when they answer a buying question, through its answer gaps and cited sources. Ask Claude to list the publishers behind those citations, then use Outreach to check whether you already have an account or a past conversation with any of them.

Treat this as research, not a send list. Claude can match cited domains to accounts already in Outreach and flag where a relationship exists. A person on your team decides who to contact and what to say. Linkeddit supplies the citation data. It does not collect contact details, and it is not a lead tool.

Check the licence, the client and the permissions. The help center says Claude is available for both Claude Desktop and the web app, and that Outreach applies the profile assigned to your user. It does not support clients that authenticate without a user, so an agent needs a licensed Outreach user identity.

Then check which Outreach instance you sign in to. Outreach says the client connects to whichever instance your credentials belong to, and recommends one instance at a time per session. Most clients open your default browser, so sign in to the right account there first.

Last, know what Claude cannot do alone. The Carly review of the Claude route notes the connector has no event triggers: nothing fires when a prospect replies, and Claude acts when you ask. A weekly competitor refresh is a saved prompt you run, not an automatic job.