Pilot input and return contract.

This pilot contract defines a closed-loop handoff between a partner and AI Security LLC: the partner submits one sanitized finding, trace, chain, or webhook object, optionally paired with system and authority context, along with the intended return location in the product and the definition of meaningful uplift. AI Security LLC then proves the signal by mapping evidence into an explicit contract, comparing attack-path behavior, qualifying the path, and enriching the authority and system graph. The process includes independent challenge and validation so the result is not just summarized, but operationally defended. The output is projected back into the partner’s native workflow as an enriched attack-path result with evidence and provenance mapping, cross-chain remediation chokepoints, retest and lifecycle fields, and a documented integration path. The pilot is complete only when the returned artifact lands where the product can use it directly, turning security analysis into an actionable product-native result rather than a detached report.