← Back to issue16 / 18 · Week of Aug 3, 2026

CRM makes evidence a write gate

The open-source CRM runs a durable research agent from a leased work queue; its tools record observed evidence, while weak matches stay as human-settled suggestions. Why it matters: A CRM agent can keep working after the browser closes. The useful boundary is not the chat surface but the rule that stops a model from turning a plausible guess into customer data.

Try this: Read SECURITY.md before using real contacts, then run the quick start with sample data and inspect the allow-list, sandbox egress, and record-write rules.

Source
GitHub — trycompai/crm
View source →

Get the field brief every week.

Important AI developments, useful explanations, and practical resources in one weekly read. Context to understand what matters, with links to the original sources and deeper reading.

Subscribe free →
Free weekly·No spam·Unsubscribe anytime