run-accent-testing

Create accent-specific personas and launch synchronized voice agent evaluation runs.

2|Updated Feb 17, 2026
One-click install
npx skills add https://github.com/coval-ai/coval-external-skills --skill run-accent-testing
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: run-accent-testing
Source: https://github.com/coval-ai/coval-external-skills/tree/main/skills/runs/run-accent-testing
Command: npx skills add https://github.com/coval-ai/coval-external-skills --skill run-accent-testing

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq.

What problem does it solve?

This skill removes the manual burden of creating multiple personas, launching individual runs, and aggregating results when testing how well a voice agent understands different accents.

Core Features & Use Cases

  • Automated Accent Sweep: Automatically creates a suite of accent-specific personas and launches parallel evaluation runs for each.
  • Comparison Reporting: Generates a consolidated comparison table and a saved multi-run report grouped by persona for immediate analysis.
  • Use Case: Use this when you need to verify that your voice agent maintains consistent performance across Indian, German, Chinese, US Southern, Nigerian, and Malaysian accents compared to a neutral baseline.

Quick Start

Use the run-accent-testing skill to launch an accent sweep for the agent named VoiceBot using the test set named Production-Set.

Frequently Asked Questions about run-accent-testing

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate voice agent accent testing across multiple linguistic profiles?▼

Voice agent accent testing is automated by creating persona packs and launching synchronized evaluation runs. This generates a consolidated comparison table and multi-run report grouped by persona for immediate analysis.

What accents can I test for speech recognition stress testing in my voice AI?▼

Speech recognition stress testing supports diverse linguistic profiles including Indian, German, Chinese, US Southern, Nigerian, and Malaysian accents, evaluated against a neutral baseline to verify consistent agent performance.

Do I need the Coval CLI and API key to run automated accent evaluation sweeps?▼

Yes, automated accent evaluation sweeps require the Coval CLI and an authenticated API key to manage persona creation, run orchestration, and report generation for voice-based AI workflows.

How do I generate comparison reports for multi-run voice agent evaluations?▼

Comparison reports for multi-run voice agent evaluations are generated automatically after launching parallel accent sweeps. The system outputs a consolidated comparison table and a saved report grouped by persona.

Does the run-accent-testing skill support parallel evaluation runs for different personas?▼

Yes, the skill automatically creates a suite of accent-specific personas and launches parallel evaluation runs for each, removing the manual burden of individual run orchestration and result aggregation.

Do I need jq installed to automate end-to-end voice agent accent testing?▼

Yes, jq is required as a dependency to automate end-to-end voice agent accent testing, supporting the processing logic for persona creation, run orchestration, and report generation.