Agentosaurus

Agentosaurus builds recursive superintelligence on your business: AI that improves itself on your own problems, on compute you control, under rules you can read back. Work starts with a fixed-scope Audit (2–4 weeks, priced after a scoping call, /agents/audit.txt) that maps what a client’s agents already change and sets the rules self-improvement runs under. It continues with LOGOS retainers on the client’s own problems, and at the top sells the stages of a sovereign training run: curated training data, supervised pretraining, fine-tuning and evaluation. For teams on Claude Code, the policy gateway (/policy-gateway.txt): A written policy for what a company’s Claude Code agents may do, enforced on every tool call through managed hooks, with a human approvals queue, a tamper-evident decision log and a weekly report per team on spend and use. Starter is free. Team, Business and Sovereign are priced after a scoping call. Buyers may be humans or autonomous agents.