Helps the user with CatchIntent platform questions — from listener setup and intent tuning through CRM integrations, MCP server, webhooks, and lead enrichment.
Step 1 — Gather context
If references/learnings.md exists, read it first for accumulated platform knowledge.
Ask the user:
-
What area of CatchIntent do you need help with?
- A) Listeners — creating, tuning keywords, relevance thresholds
- B) Signals — understanding scores, filtering, managing quota
- C) AI response suggestions — improving quality, customizing tone
- D) CRM integrations — HubSpot, Pipedrive, Close, Lemlist, Apollo, Instantly
- E) Webhooks — setting up, payload issues, debugging
- F) MCP server — connecting to Claude, Cursor, or ChatGPT
- G) LinkedIn Intelligence — Agentic Search, buyer identification
- H) Alerts — Slack, Discord, Telegram, email notification setup
- I) Team & billing — plan features, quota management, team members
- J) Something else — describe it
-
What's your goal? (describe your specific question or problem)
If the user's request already provides most of this context, skip directly to Step 2. Lead with your best-effort answer using reasonable assumptions (stated explicitly), then ask only the most critical 1-2 clarifying questions at the end.
Step 2 — Route or answer directly
If the request maps to a specialized skill, route:
- Social listening strategy or tool comparison →
/sales-social-listening [question] - Buyer intent signals and prioritization →
/sales-intent [question] - Contact enrichment →
/sales-enrich [question] - CRM selection →
/sales-crm-selection [question] - HubSpot-specific config →
/sales-hubspot [question]
Otherwise, answer directly from the platform reference below.
Step 3 — CatchIntent platform reference
Read references/platform-guide.md for the full platform reference — capabilities, pricing, data model, integration recipes, MCP setup.
Answer the user's question using only the relevant section. Don't dump the full reference.
Step 4 — Actionable guidance
You no longer need the platform guide — focus on the user's specific situation.
- Listener tuning — start threshold at 70-75%, lower if too few signals, raise if too noisy
- Quota management — prioritize high-intent keywords, remove broad terms eating quota
- CRM sync — verify plan supports CRM integrations (Pro+ only), check field mapping
- MCP setup — use OAuth flow, no API key needed
- Alert routing — match channels to urgency (Slack for immediate, email for digests)
If you discover a gotcha, workaround, or tip not covered in references/learnings.md, append it there.
Gotchas
Best-effort from research — review these, especially items about plan-gated features and integration gotchas that may be outdated.
- CRM integrations are Pro+ only. Basic plan ($49/mo) gets monitoring and alerts but no CRM push. Upgrade to Pro ($69/mo) for HubSpot, Pipedrive, Close integrations.
- Webhooks are Pro+ only. Basic plan supports email and Slack alerts but not webhooks.
- Signal quota is hard-capped. 150 signals/mo on Basic, 500 on Pro. No rollover. Broad keywords can eat quota fast — tune relevance thresholds aggressively.
- No traditional REST API. CatchIntent exposes an MCP server (OAuth-based) rather than a REST API. Use the MCP server for programmatic access from Claude, Cursor, or ChatGPT.
- LinkedIn Intelligence is a separate product. Starts at $69/mo on top of social listening. Not included in Basic or Pro social listening plans.
- AI response suggestions are Pro+ only. Basic plan surfaces signals but doesn't draft replies.
- 7-day trial requires card on file. Cancel before day 7 to avoid charges. 14-day refund policy after that.
Related skills
/sales-social-listening— Social listening strategy — monitoring setup, Boolean queries, sentiment analysis, competitive intelligence, crisis detection, tool comparison/sales-octolens— Octolens — developer-first social listening with MCP server, REST API, webhooks across Reddit, GitHub, HN, X, LinkedIn, Bluesky/sales-syften— Syften — AI-filtered keyword monitoring across Reddit, HN, X, Bluesky, 15+ platforms, sub-minute alerts, REST API, webhooks/sales-brand24— Brand24 — social listening with MCP server, sentiment analysis, Share of Voice/sales-reddinbox— Reddinbox — AI audience intelligence with intent scoring across Reddit, X, Bluesky, HN/sales-intent— Buyer intent signals and prioritization from monitoring data/sales-enrich— Contact enrichment strategy across tools/sales-do— Not sure which skill to use? The router matches any sales objective to the right skill. Install:npx skills add sales-skills/sales --skill sales-do
Examples
Example 1: Set up multi-platform monitoring
User says: "I want to monitor Reddit and Hacker News for people looking for a project management tool like mine" Skill does:
- Creates a listener with keywords matching project management pain language
- Sets relevance threshold to 70% for broad initial coverage
- Configures Slack alerts for immediate notification
- Recommends reviewing first 20 signals and adjusting threshold based on noise level Result: Listener running with tuned intent detection across both platforms
Example 2: Connect CatchIntent to HubSpot CRM
User says: "How do I push CatchIntent leads into my HubSpot pipeline automatically?" Skill does:
- Confirms user is on Pro plan (CRM integrations require Pro+)
- Walks through HubSpot native integration setup
- Maps CatchIntent signal fields to HubSpot contact/deal properties
- Sets up webhook as backup for custom field mapping Result: Qualified signals automatically create contacts in HubSpot with intent context
Example 3: MCP server integration
User says: "I want to use CatchIntent from Claude Code to check my latest signals" Skill does:
- Points to the CatchIntent MCP server (OAuth-based, no API key needed)
- Walks through connecting the MCP server in Claude settings
- Shows example natural language queries: "Show me today's high-intent signals" or "Find Reddit threads about CRM alternatives" Result: CatchIntent data accessible directly from Claude via MCP
Troubleshooting
Listeners returning too much noise
Symptom: Most signals are irrelevant — casual mentions, jokes, or unrelated discussions Cause: Relevance threshold set too low or keywords too broad Solution: Raise the relevance threshold from default to 80-85%. Remove generic keywords and use more specific phrases that match buyer language ("looking for", "need a tool", "alternative to"). Review the top 20 signals to identify patterns in false positives.
Signal quota exhausted mid-month
Symptom: "Signal limit reached" before your billing cycle resets Cause: Broad keywords generating too many low-value signals Solution: Audit your listeners — remove keywords that generate high volume but low relevance. Increase the relevance threshold to only capture high-intent signals. Consider upgrading from Basic (150) to Pro (500) if you genuinely need more coverage.
CRM integration not syncing leads
Symptom: Signals appear in CatchIntent dashboard but don't show up in HubSpot/Pipedrive Cause: Plan doesn't support CRM integrations, or integration not properly authorized Solution: Verify you're on Pro plan or higher (Basic doesn't include CRM integrations). Re-authorize the CRM connection in Settings > Integrations. Check that the target pipeline/list exists in your CRM. Test with a single manual push before enabling auto-sync.