mcpField
S
support-quality-qa-review

support-quality-qa-review

github.com/sylphxaiTesting & quality145 installsUpdated today
Repository

About support-quality-qa-review

A skill for support and QA teams to assess customer-service quality and produce a scorecard, review workflow, event schema, and improvement backlog. It emphasizes correctness, fairness, and trust alongside operational performance.

Key features

  1. 01Define support channels and risk classes
  2. 02Set QA dimensions and scoring criteria
  3. 03Design sampling and calibration workflows
  4. 04Review macros and automation
  5. 05Create event schema and improvement backlog

Use cases

  1. 01Building a support QA scorecard
  2. 02Planning agent coaching and calibration
  3. 03Reviewing automated high-impact responses

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/sylphxai/skills --skill support-quality-qa-review

After installing, type/support-quality-qa-reviewin Claude Code to confirm it loaded.

Bundled filesagents/openai.yamlreferences/support-quality-patterns.md

Details

From the project README

Support Quality QA Review

Use this skill to convert a risky product, operations, trust, or marketplace question into a concrete artifact with owners, gates, metrics, and recovery paths.

Workflow

  1. Identify support channels, case types, risk classes, macros, escalation paths, QA owner, and customer outcomes.
  2. Read references/support-quality-patterns.md.
  3. Define QA dimensions: accuracy, empathy, policy, evidence, resolution, escalation, safety, revenue/trust impact, and documentation.
  4. Design sampling, calibration, coaching, macro improvement, automation review, and product feedback loops.
  5. Produce QA scorecard, review workflow, event schema, and improvement backlog.

When not to use

  • Do not use when the job belongs to customer-support-operations — Defer when the job matches customer-support-operations instead.
  • Do not use for generic advice the base model already handles without this skill's specific artifact contract.

Guardrails

  • Do not score agents only on speed when correctness, fairness, or trust is at risk.
  • Do not punish support for product gaps; route recurring causes back to product owners.
  • Do not let AI/macros send high-impact policy, billing, or safety responses without QA gates.
  • Protect customer privacy in QA examples and coaching artifacts.

Output format

Support QA context:
Channel / case type / risk class:

Scorecard:
| Dimension | Good | Risk | Weight | Evidence | Coaching action |
| --- | --- | --- | --- | --- | --- |

QA workflow:
- <sample/review/calibrate/coach/improve>

Product feedback:
- <recurring issue> -> <owner/artifact/metric>

Comments

Markdown supported