Claims and boundaries
What TrustGate Sovereign describes—and what it does not claim
TrustGate Sovereign connects business authority, protected execution and evidence so teams can understand what was authorized, what happened and what remains unresolved. Public evidence includes a real Salesforce Developer Edition test for a specifically assessed action path; customer deployment is qualified separately against its own requirements.
Publicly described principles
- Exact-action authorization evaluates one consequential action rather than treating a governed system as authority for every action.
- Accountable enterprise authority is bound to the action.
- The protected target and permitted route are part of the authorization boundary.
- Target-authoritative evidence, not the initiator's assertion alone, establishes the external outcome.
- Unknown and contradictory outcomes remain explicit rather than being represented as success.
- Reconciliation keeps an unresolved outcome open while further authoritative target evidence is requested.
- A retry, reversal, refund, correction or recovery is a separately authorized corrective action; it requires new authority and a new clearance.
- Outcome proof preserves reviewable evidence of what the protected target accepted, rejected or left unresolved where the bounded path can show it.
- Public evidence includes bounded controlled scenarios and the real-target Salesforce consequence-control proof.
- TG360 presents a buyer evidence posture without becoming a second decision engine.
- Adoption proceeds through scoped assessment and a bounded pilot path.
Controlled private review
A controlled private review may cover:
- Accepted and rejected synthetic scenarios.
- Architecture and target-pattern assessment.
- Richer TG360 evidence.
- Claims-boundary review.
- DORA-oriented and data-protection evidence review without a compliance or legal conclusion.
- Target-specific integration context.
Evidence and adoption limits
- This public site does not claim a completed production customer deployment.
- No readiness for every target.
- No universal zero-code integration.
- No guaranteed outcome.
- No enterprise-wide non-bypassability.
- No legal, regulatory, compliance or security certification.
- No partner or vendor endorsement.
- No replacement of human review.
- No replacement of customer acceptance.
- No public disclosure of proof semantics, private contracts, credentials, topology, source code or bypass-relevant internals.
Beyond preventing an unsafe retry
Consequential operation explainability
Show what was authorized, what was executed, what remains uncertain and which evidence justified a continuation or closure decision. This supports operational investigation and accountability.
Consequence-control economics
Reuse a consistent authority, execution and evidence-control pattern across qualified agent workflows, rather than rebuilding equivalent controls for each one. Evaluate potential savings, deployment speed and ongoing maintenance costs against the customer's existing approach.
Maturity and adoption boundary
An initial review can use bounded browser scenarios and the Salesforce Developer Edition reference, without customer production data. A customer pilot requires an agreed scope, target-specific assessment, and customer-specific integration and implementation work.
Production operation is considered only after customer acceptance of the agreed design and controls. Each corrective action requires fresh accountable authority and a new clearance; an earlier clearance is not silently reused.