How to Manage Your Sales Signals from Claude Code

What if your morning sales routine was a conversation?

“Show me today’s hot signals on Reddit. Draft a casual reply for the top 3. Enrich the people behind them and push to HubSpot.”

That’s not a hypothetical. It’s what CatchIntent + Claude Code looks like with MCP.

TL;DR: CatchIntent now works as an MCP server. Connect it to Claude Code (or Cursor, Windsurf, ChatGPT) with one command. Get access to 20 tools — search signals, manage listeners, draft AI outreach, enrich prospects, export CSVs, push to CRM. All through natural language. Available on all plans.

What You Need

A CatchIntent account (free trial works)
Claude Code installed

That’s it.

Setup (30 Seconds)

Open your terminal and run:

claude mcp add catchintent --transport http https://engine.catchintent.com/mcp

Start a new Claude Code session. The first time you use a CatchIntent tool, your browser opens for a one-time OAuth authorization. Pick your workspace, click Authorize, done.

The Morning Signal Review

Every GTM person starts their day the same way: check what came in overnight, prioritize, act.

You say:

“Check my CatchIntent signals from today. Focus on hot and warm ones.”

Claude searches your signals, filters by warmth, and returns a summary. You see platform, intent type, relevance score, and the key extracted data (budget mentioned, competitors, pain points) — all in your terminal.

Then:

“Draft a casual reply for the first two signals.”

Claude calls generate_response for each signal. It uses your brand voice configuration and the full signal context (the original post, AI reasoning, extracted data) to draft a response. You review, tweak if needed, or ask it to regenerate with feedback:

“Make the second one shorter and less salesy.”

Finally:

“Mark both as reached_out.”

Done. Two signals reviewed, two responses drafted, two statuses updated. No dashboard opened.

Prospect Enrichment Batch

Your SDR needs emails for this week’s outreach. Normally that’s: open dashboard → filter people → click enrich one by one → export.

With MCP:

“Find all approved prospects with strong ICP match who don’t have emails yet. Enrich the top 10.”

Claude searches your people, filters by ICP and enrichment status, triggers enrichment for each. You get back names, titles, companies, and verified emails.

“Export them as CSV.”

A CSV drops into your conversation. Copy it, send it to your SDR, done.

Listener Management

Your “Mailchimp alternatives” listener has a 60% AI rejection rate — too many irrelevant posts. You could dig through the dashboard to find which keywords are underperforming. Or:

“Check the health of my Mailchimp listener. Which keywords have the worst performance?”

Claude calls get_listener with health metrics, identifies the problem keywords, and suggests changes.

“Remove ‘email’ and add ‘email marketing platform’. Lower the threshold to 65.”

One update_listener call later, your listener is tuned.

Creating a New Listener

A competitor just launched. Time to monitor the backlash.

“Create a listener called ‘CompetitorX complaints’ tracking ‘CompetitorX sucks’, ‘CompetitorX alternative’, ‘switching from CompetitorX’ on Reddit and X. Set purpose to competitive intelligence.”

Claude creates the listener, confirms the keywords and platforms, and it starts monitoring immediately.

Pipeline Cleanup

Friday afternoon. Time to clean up the pipeline before the weekend.

“Show me my signal board. How many are stuck in reached_out with no response? And what’s my win rate this month?”

Claude pulls the kanban view, identifies stale signals, and shows your dashboard stats. You make decisions:

“Mark these three old no-response signals as lost. And push signal X to my CRM — that one’s looking good.”

What’s Available

20 tools in total:

Category	Tools
Signals	Search, get details, board view, dropped signals, update status, generate response
Listeners	List, get details, create, update, toggle
People	Search, get details, update status, enrich, export CSV, search companies
CRM	Push to HubSpot/Close/Pipedrive
Workspace	Dashboard stats, usage

All operations respect your plan limits and role permissions — same rules as the dashboard.

Pre-Built Skills

Don’t want to prompt from scratch? Install our GTM skills:

npx skills add catchintent/skills

Six workflows ready to go:

Morning Signals — daily review with prioritized actions
Draft Outreach — AI responses in your brand voice
Enrich Prospects — batch enrichment + export
Pipeline Review — pipeline health + follow-up recommendations
Listener Tune — keyword performance analysis
Weekly Report — team-ready summary

Each skill is a playbook that tells Claude how to use CatchIntent tools effectively. You just say what you want.

Why This Matters

The dashboard isn’t going away. It’s still the best way to onboard, manage your team, and get a visual overview.

But for the daily grind — reviewing signals, drafting responses, enriching prospects, tuning listeners — natural language is faster. You describe the outcome, the AI handles the workflow.

That’s what AI-native means. Not AI as a feature. AI as the interface.

Get started: Sign up for CatchIntent → Run the setup command → Start prompting.

Already a user? Go to Settings > Agents in your dashboard, or just run:

claude mcp add catchintent --transport http https://engine.catchintent.com/mcp