About Sigillum

Sigillum gives an outside verification layer to software actions that are easy to automate and expensive to get wrong. Instead of trusting the same system that generated the change to approve it, Sigillum sells a neutral receipt with a score, recommendation, findings, payment proof, and agent decision.

The dashboard is the proof window. The real product is the inspection contract: quote first, pay through x402, inspect the risky action, and return a receipt that tells the caller whether to continue, pause, or block.

AI tools can now generate code changes, dependency installs, and deploy actions in seconds. The risk is not just incorrect code; it is the lack of a neutral checkpoint before production-impacting behavior happens.

Sigillum exists to create that checkpoint with a public-safe proof surface and a repeatable machine contract.

• Neutral review before merge, install, or deploy
• Paid verification rather than silent internal trust
• Persisted receipts and lifecycle proof instead of untraceable approvals

Every serious Sigillum surface should preserve the same loop, whether the caller is a CLI user, a GitHub maintainer, or an outside autonomous agent.

• Request quote
• Receive HTTP 402 payment requirement
• Pay through x402
• Run inspect
• Return receipt and agent decision
• Continue, pause, or block based on the result