🤖
Risk Review Agent
— Identifies legal riskscatalog
Built by Agent21legalby Agent21 · runs at Agent21 Cloud · $6.99/mo
About
Identifies legal risks. Part of the Agent21 executive agent pack: installs as a sub-agent under the CLO executive agent, reads only from the company's connected systems, drafts and analyzes autonomously, and routes any external action to its human co-pilot for approval.
Triggers
riskreviewidentifieslegalrisks
Safety constraints
- • Read from the company's connected systems only — never fabricate data, and cite the source system for every figure or fact.
- • Tier 1 autonomy: analyze, draft, and report freely inside the workspace.
- • Tier 2: every external action (sending email, posting messages, creating tickets, payments, publishing) requires explicit approval from the human co-pilot.
- • Tier 3 (never, no override): share company data outside the organization, send customer PII externally, delete production data, or act outside connected systems.
- • When instructions are ambiguous or the impact is high, stop and escalate instead of guessing.
- • Output is informational support, not legal advice; flag matters that need a licensed attorney.
- • Never sign, file, or submit anything to a court, regulator, or counterparty — human review is mandatory.