Exact-version qualification
Registration, evidence binding, and version-specific findings across the agreed agent scope.
GhostGate AI Agent Qualification Pilot
A bounded six-week engagement for consequential AI-agent releases.
For organizations preparing consequential AI agents for production and needing broader qualification across multiple agents, versions, workspaces, policies, and deployment workflows.
Standard engagement
Six-week qualification pilot
$32,000 USD01 / Standard scope
A defined qualification boundary.
The standard pilot constrains the organization, workspaces, versions, policies, scenarios, and working sessions so the engagement can reach an evidence-backed release decision in six weeks.
- OrganizationONE
- WorkspacesUP TO TWO
- AgentsUP TO FIVE
- Exact versionsUP TO TEN
- Policy packsTWO
- Agreed scenariosFORTY
- Working and review sessionsFIVE
- Engagement durationSIX WEEKS
02 / Decision package
Technical evidence with executive meaning.
The result is a defensible decision on which exact AI-agent versions may proceed, under what restrictions, with what evidence, and when requalification is required.
- Qualification findings
- Behavior and drift results
- RiskChain and dependency findings
- Release decisions and conditional controls
- Signed deployment attestations
- Sanitized evidence archive
- Technical and executive closeout
03 / What the price covers
Software plus senior qualification judgment.
The pilot combines repeatable GhostGate software with the human work required to make qualification relevant to the customer’s actual release boundary.
Senior security judgment
Interpretation of behavior, permissions, causality, fleet relationships, and material change.
Policy calibration
Two policy packs aligned to the customer’s authority, risk, and human-approval requirements.
Scenario design
Forty agreed scenarios focused on the consequential release questions in scope.
Fleet and dependency analysis
Cross-agent relationships, shared identity, shared dependency, and potential outbreak clusters.
Signed evidence generation
Version-bound attestations and a sanitized archive suitable for technical and governance review.
Technical workshops
Working sessions to establish scope, calibrate policy, inspect findings, and make release decisions.
Executive interpretation
A closeout that translates technical evidence into production conditions and requalification triggers.
Founder-led delivery
Direct delivery across qualification design, review, evidence, and closeout.
04 / Commercial terms
Clear boundaries before kickoff.
- Duration: six weeks.
- Payment: 50 percent at kickoff and 50 percent at delivery.
- Scope control: additional agents, versions, scenarios, integrations, and workshops require a scope change.
- Data boundary: the standard pilot does not require production secrets and can use sanitized fixtures and bounded environments.
- Release authority: the customer retains the final production decision.
Expanded qualification
Determine whether the private pilot fits your release program.
The first conversation is a 20-minute technical review. Teams with one exact release decision can begin with the fixed Release Check.