What it does
Intercom Churn Signal Detector helps customer success teams identify at-risk accounts before they churn. It fuses Intercom conversation signals (volume, sentiment, repeat issues over the last 60 days) with Stripe billing events (payment failures, pauses, downgrades over the last 90 days) into a deterministic churn risk score, then routes high-risk accounts to Salesforce as opportunities and alerts the CS team via Slack. A single-signal guardrail prevents false positives: accounts with only one category of risk signal go to a watch list, not a CRM opportunity. Churn risk scores are strictly internal — never shared with customers. Built for customer success and RevOps teams who want early, explainable churn detection instead of reactive account loss.
Example prompts
- Run the full churn detection pipeline and alert Slack for high-risk accounts
- Which accounts need immediate CSM action right now?
- Scan this account's churn risk signals
- Generate a CSM alert for the highest-risk accounts this week
Tools (8)
Tools the agent exposes — your AI client calls them automatically when it needs them.
- run_full — Run the full churn-signal detection pipeline: collect Intercom + Stripe data, score churn risk, generate CSM narrative, return the complete ChurnAgentResult JSON.
- analyze_churn_risk — Analyze pre-collected RawData (Intercom + Stripe) into scored ChurnFinding[]. Pure deterministic analysis, no LLM.
- get_high_risk_accounts — Filter to accounts requiring immediate CSM action (risk_level=high), with Salesforce opportunity recommendation and at-risk MRR.
- generate_csm_alert — Generate a Slack-ready CSM alert message for the highest-risk accounts. Internal use only.
- scan_account_signals — Inspect a single account's churn risk signals without running the full pipeline.
- discover_intent — Intent-discovery front door: state a freeform goal and get a co-designed input proposal.
- plan_inputs — Return questions, schema and a ready-to-edit example for a named tool.
- list_capabilities — List agent capabilities: tools, scoring formula, credential slots, and guardrails.
What you'll need to connect
This agent will ask you for the following. You enter them when you connect — they're encrypted and never shared with the creator.
- Intercom Access Token · optionalIntercom API access token with conversations:read and contacts:read scopes. Leave blank to use mock fixture data.Get this from intercom.io's account or API settings.Paste the value as a single line.Only sent to: api.intercom.io
- Stripe Secret Key · optionalStripe API key with subscriptions:read, invoices:read, events:read scopes. Leave blank to use mock fixture data.Copy a secret key from the Stripe Dashboard (Developers → API keys).Paste the value as a single line.Only sent to: api.stripe.com
- Salesforce Access Token · optionalSalesforce OAuth access token with Opportunities:write permission, used to create churn-risk opportunities.Get this from login.salesforce.com's account or API settings.Paste the value as a single line.Only sent to: login.salesforce.com, *.salesforce.com
- Salesforce Instance URL · optionalSalesforce instance URL, e.g. https://yourorg.my.salesforce.com.Get this from *.salesforce.com's account or API settings.Paste the value as a single line.Only sent to: *.salesforce.com
- Notion API Key · optionalNotion API key with pages:write permission for the monthly churn summary.Create an internal integration in Notion (notion.so/my-integrations) and copy its secret, then share the relevant pages with that integration.Paste the value as a single line.Only sent to: api.notion.com
- Slack Bot Token · optionalSlack bot token for posting to a #cs-alerts channel.Create a Slack app (api.slack.com/apps), add the scopes the agent needs, install it to your workspace, and copy the bot/user token.Paste the value as a single line.Only sent to: slack.com
- Anthropic API Key · optionalOptional. Leave blank on a sampling-capable client to use the host model with no key.Create a key on the Anthropic Console API keys page (console.anthropic.com → API keys).Paste the value as a single line.Only sent to: api.anthropic.com
How you're protected
FindAgent runs these safety checks on every agent automatically. They're always on and can't be turned off.
- Prompt-injection scanning
Every request is checked for known prompt-injection and jailbreak attempts before the agent runs. This is always on.
- Secret-leak scanning
Every response is scanned for leaked API keys, tokens, and other secrets before it reaches you. This is always on.