Sample TrustScore Evidence Pack
This is the artifact a buyer, assessor or procurement team receives — a cryptographically sealed record of the measurement, the admissibility verdict, and the exact engine that produced them. Below is a redacted example for a fictional project.
Adequate trust signal; investment-grade range.
Signal only. The TrustScore measures posture — it does not, on its own, authorize a release.
Blocked — critical security gate. kev_block must clear; score is not a factor.
The verdict enforces policy gates. A strong score never overrides a failed security-hard gate.
TrustScore measures posture. Verdict enforces policy gates.
That is why this project can sit in the investment-grade BBB band and still be blocked: a single known-exploited vulnerability fails a non-waivable security gate, and no score can clear it.
Admissibility gates
Every gate is evaluated to a pass / warn / fail predicate. Class precedence is fixed: a security-hard failure dominates the verdict regardless of lower-class gates.
| Gate | Class | Result | Waiver | Detail |
|---|---|---|---|---|
| kev_block | Security (hard) | FAIL | Non-waivable | 1 known-exploited (KEV) finding — CVE-2026-1337 [cisa-kev-2026-06-19] |
| critical_vuln | Security (hard) | PASS | Dual control | 0 unwaived critical findings (max 0) |
| pass_threshold | Policy threshold | PASS | Non-waivable | TrustScore 79.3 vs threshold 75 |
| domain_fail | Policy threshold | PASS | Exec acceptance | All domains above fail floor (>5) |
| pqc_floor | Policy threshold | PASS | Profile | pqc_readiness signal 61.0 (floor proxy 5) |
| domain_warn | Advisory | WARN | n/a | ai_risk domain in warn band (18.4 < 20) |
Score breakdown — six weighted domains
Signed manifest & integrity seal
The verdict, the engine constants and the version tuple are sealed inside a SHA-256 (optionally HMAC) digest, so the decision a buyer acts on is attested — not re-derived outside the audit path.
{
"scoring_engine_version": "0.2.0",
"policy_version": "2.4",
"decision_policy_version": "1.0",
"decision_policy": {
"pass_threshold": 75,
"critical_vuln_max": 0,
"kev_feed_version": "cisa-kev-2026-06-19"
},
"verdict": {
"decision": "block_security",
"authoritative_gate_class": "security_hard",
"failed_blocking_gates": ["kev_block"],
"deployable": false,
"gate_result": "fail",
"measurement": { "trust_score": 79.3, "band": "BBB", "tier": "Lower Medium" }
},
"engine_constants": { "ess_basis": "pre_clip", "band_window": 2 },
"seal": {
"algorithm": "SHA-256",
"digest": "9f2c41a7e0b8d63f5a1c84be72d09f1c6a4e3b8d05f7c29a1e4b6d8f03a2c7e91"
}
}Digest shown is illustrative. In a live pack the seal verifies against the sealed payload and rejects any engine-constant divergence or non-canonical version.
Bring evidence, not assertions, to your next review
Generate signed evidence packs your assessors and procurement teams can verify independently.