HuskyVoiceAI

Are scope, spend, and authority enforced while the agent runs?

Rubric 0 as written -- no enforced limits are published. runtime_governance and permission_scopes are both no_public_information against a capture read in full. The page does publish one real control, escalation of complex or urgent calls to staff with full context as warm transfers rather than cold drops, and that is credited in human_oversight. It is not scored here because a fallback that hands a call to a person is not a scope, spend or authority limit enforced at run time, and no threshold, cap or boundary of any kind is stated. Compare pally (2026-10-12), which earned 2 on concrete self-attested caps -- a single-use card limited to the approved amount. There is no equivalent here.

Rests on Human oversight, each cited below. Also named Runtime governance and Permission scopes, which our record now shows as undisclosed; that part of the support no longer holds. What a score rests on

Is there a tamper-evident record of what it actually did?

Every call is summarised and pushed to the customer's CRM, Sheets, Slack or business systems with sentiment, intent and next actions, so a per-call record does reach systems the customer controls. Held to 1 rather than 2 because what is published is a business summary of the conversation, not a log of what the agent did or was permitted to do, and audit_trail_immutable is no_public_information, so nothing claims tamper-evidence. A reader cannot reconstruct or verify the agent's own actions from it.

Counts Audit trail, each cited below. Also named Tamper-evident log, which our record now shows as undisclosed; that part of the support no longer holds. What a score rests on

Are compliance obligations attached per engagement?

No public evidence found. compliance_certs and regulatory_alignment are both no_public_information against a fully read capture. This is a material absence rather than a neutral one: the same page markets financial-services work (card enquiries, loan follow-ups, payment reminders, verification calls) as handled "securely and at scale", and healthcare, dental, diagnostics and hospital booking, and publishes no certification, audit, or regulatory alignment to support operating in either sector.

Rests on High-risk domains, each cited below. Also named Compliance certifications and Regulatory alignment, which our record now shows as undisclosed; that part of the support no longer holds. What a score rests on

Does payment depend on a verified result?

No public evidence found. settlement_mechanism and pricing_model are both no_public_information against a fully read capture -- the page publishes no pricing at all, so there is nothing linking payment to a verified outcome, and equally no evidence of per-seat or per-token billing. Scored 0 as an absence finding, not as a judgement that billing is result-independent.

Rests on Settlement mechanism and Pricing model, recorded as undisclosed and cited below. For a 0 that absence is the finding. What a score rests on

Can the deployment be insured, and is the customer indemnified?

No public evidence found. insurance_available, insurance_carriers, indemnification and coverage_limits are all no_public_information against a fully read capture. No carrier, no limit, no customer indemnity.

Rests on Insurance available, Insurance carriers, Coverage limits and Indemnification, recorded as undisclosed and cited below. For a 0 that absence is the finding. What a score rests on