dvcrn Outlit 🔦
Outlit is customer intelligence for agents: a single CLI that joins product activity, web signals, billing, and conversations into a unified customer context graph + timeline.
Choose this skill when you need cross-tool answers like:
- "What changed for this customer this week?"
- "Who's paying but inactive for 30 days?"
- "What objections show up in conversations?"
- "Which channels drive revenue?"
Quick start (CLI)
npm install -g @outlit/cli # or: brew install outlitai/tap/outlit
outlit auth login # or: outlit auth login --key ok_your_api_key_here
outlit customers get acme.com --include users,revenue
Auth lookup order (highest → lowest): --api-key, OUTLIT_API_KEY, stored config.
Output contract
- Interactive TTY → readable tables
- Piped stdout / CI → automatic JSON (no flags needed; force with
--json) - Every result includes timestamps + source attribution for traceability
Proactive churn insights
Outlit also surfaces proactive churn-risk insights by correlating signals across tools (e.g., auth failures + support email + subsequent silence), with recommended actions.
Signal extraction
Facts — AI-extracted structured signals, not raw events. Returns business-level insights like "champion left", "budget approved", or "usage dropped 40%". Time-windowed so you can ask "what changed this quarter?"
Search — Semantic natural-language search across all customer interactions and conversations. "pricing objections" matches discussions about cost concerns, budget pushback, etc. — not just keyword hits. Scope to one customer or search org-wide.
Core commands
# List customers (filters)
outlit customers list --billing-status PAYING --no-activity-in 30d
# Get a profile (limit included sections)
outlit customers get acme.com --include users,revenue,recentTimeline --timeframe 30d
# Timeline (scoped)
outlit customers timeline acme.com --timeframe 90d --channels EMAIL,SLACK --limit 50
# Natural language search (org-wide or scoped)
outlit search "pricing objections"
outlit search "budget concerns" --customer acme.com --after 2025-01-01 --before 2025-03-31
# Facts / signals
outlit facts acme.com --timeframe 30d
# Users (filter by journey stage, activity, customer)
outlit users list --journey-stage CHAMPION
outlit users list --no-activity-in 30d --journey-stage AT_RISK
# SQL analytics (read-only)
outlit schema
outlit sql "SELECT event_type, COUNT(*) FROM events GROUP BY 1"
MCP (for agents that support MCP)
Endpoint: https://mcp.outlit.ai/mcp
Auth header: Authorization: Bearer YOUR_API_KEY
Outlit MCP tools include: list customers/users, get customer/timeline/facts, search customer context, run SQL, inspect schema.
Configure AI agents fast
outlit setup --yes
This auto-detects supported agents and configures MCP where applicable. It can also set up OpenClaw specifically:
outlit setup openclaw
outlit doctor # diagnose setup issues (missing key, unconfigured agents)
Links (go deeper)
- Docs home: https://docs.outlit.ai/
- Quick Start: https://docs.outlit.ai/tracking/quickstart
- CLI Overview: https://docs.outlit.ai/cli/overview
- CLI Command Reference: https://docs.outlit.ai/cli/commands
- AI Agent Setup: https://docs.outlit.ai/cli/ai-agents
- MCP Integration: https://docs.outlit.ai/ai-integrations/mcp
- Customer Context Graph: https://docs.outlit.ai/concepts/customer-context-graph