Voice Agent Governance Kit — v1.0 Release Notes

Released: 2026-05-04

Version: 1.0

Category: Compliance & Governance

Price: $4.99 (one-time)

What is New in v1.0

First public release of Voice Agent Governance Kit.

Assess any voice-agent, contact-centre, sales dialler, or support automation deployment for governance readiness in one command
Eight control domains checked automatically: consent, disclosure, recording retention, audit logging, escalation paths, voice-clone/deepfake controls, PII redaction, and abuse monitoring
Evidence-based scanning: searches all readable policy docs, call-flow notes, JSON/YAML configs, and Markdown files in the supplied path
Regulated mode (--regulated): escalates consent, disclosure, recording retention, and escalation findings to blocking errors for contact-centre, finance, health, and public-sector deployments
Clear governance status: PASS, WARN, or FAIL with per-control evidence summary and coverage table
Launch recommendation text generated automatically based on findings
Dual output formats: markdown (human-readable governance report) and JSON (for compliance dashboards and CI gates)
Handles mixed file types: scans .md, .txt, .json, .yaml, .yml, .py, .js, .ts files in a directory in one pass
Non-destructive read-only scan: never modifies any file, config, or policy

User-supplied path only; no access to MEMORY.md, USER.md, SOUL.md, openclaw.json, or fixed sensitive paths
No subprocess execution, no network calls, no eval/exec, no shell access
stdlib-only (argparse, json, pathlib — no pip installs required)
Finding evidence is keyword-based from supplied documents only — no external lookups

# Markdown governance report (default)

python3 scripts/voice_governance_audit.py /path/to/policy-docs

# JSON output for compliance dashboards

python3 scripts/voice_governance_audit.py /path/to/policy-docs --format json

# Regulated mode — contact-centre, finance, health, public-sector

python3 scripts/voice_governance_audit.py /path/to/policy-docs --regulated

# Regulated + JSON for automated CI gating

python3 scripts/voice_governance_audit.py /path/to/policy-docs --regulated --format json

PASS: All eight control domains have evidence coverage
WARN: One or more controls are missing evidence (acceptable with risk acceptance outside regulated contexts)
FAIL: One or more blocking controls are missing (regulated mode) or no readable files found

Exit status: 0 for PASS/WARN, 1 for FAIL.

| Control | What is Checked |

|---------|----------------|

| Consent | opt-in, permission, consent language |

| Disclosure | "AI agent", "automated", "not human" disclosure text |

| Recording Retention | retention policy, delete-after period, storage policy |

| Audit Logging | audit log, trace, transcript ID references |

| Escalation | human handoff, supervisor transfer, escalation path |

| Voice Clone Control | voice clone policy, watermark, deepfake, voice consent |

| PII Redaction | PII redaction, masking, personal data handling |

| Abuse Monitoring | abuse detection, fraud, misuse, rate limiting |