Quivly Skills

Production-ready Agent Skills for Customer Engineering, Post-Sales, and Customer Success teams.

Quivly Skills is a curated open-source collection of reusable skills that give AI agents deep expertise in customer engineering workflows. Every skill follows the official Agent Skills specification.

Topics: agent-skills ai-agents customer-success post-sales customer-engineering anthropic claude skills ai-workflows

About

This repository provides production-ready skills focused on real customer engineering and post-sales use cases, including health reviews, expansion playbooks, churn risk detection, QBR preparation, and technical workflows.

Why Quivly Skills?

Customer engineering and post-sales work involves repeatable but high-context processes: health reviews, expansion planning, churn mitigation, technical onboarding, QBR preparation, and more.

Quivly Skills turn that expertise into portable, version-controlled skills that any Agent Skills-compatible agent can discover and use.

• Follows the open standard — Compatible with the Agent Skills specification (originally from Anthropic, now widely adopted).
• Designed for real teams — Built from patterns that work at scale in B2B post-sales and customer engineering organizations.
• Progressive disclosure — Agents load just the metadata first, then full instructions only when needed.
• Composable — Use individual skills or combine them for complex workflows.

Quick Start

Using in Claude Code

/plugin marketplace add quivly/skills

Then install skill sets or individual skills.

Manual use (any compatible agent)

1. Clone or copy the skill directory you need.
2. Place it in your agent's skills folder (exact location depends on the client).
3. The agent will discover the skill via its name and description.

See the Agent Skills documentation for client-specific instructions.

Repository Structure

├── customer-success/     # Health scoring, churn signals, QBRs
├── expansion/            # Playbooks for growth and upsell
├── technical-onboarding/ # Implementation and technical workflows
├── communications/       # Executive updates, stakeholder updates
├── analysis/             # Data interpretation and reporting
└── template/             # Starter template for new skills

Each skill is a self-contained folder with a SKILL.md file (plus optional scripts/, references/, assets/).

Creating a New Skill

Use the template:

cp -r template/my-new-skill ./category/my-new-skill

Edit SKILL.md:

---
name: my-new-skill
description: Clear description of what this skill does and when an agent should use it. Include trigger keywords.
license: MIT
metadata:
  author: quivly
  version: "1.0"
---

# My New Skill

## When to Use
...

## Steps
1. ...

See the full specification for all supported frontmatter fields.

Contributing

We welcome contributions from Customer Engineering and post-sales practitioners.

1. Follow the Agent Skills format strictly.
2. Write excellent description fields — this is how agents decide when to activate the skill.
3. Keep SKILL.md focused. Move detailed references to references/.
4. Test the skill with at least one compatible agent.
5. Open a PR with a clear description of the use case.

Please read the skills in this repo for style and structure examples.

License

This repository is licensed under the MIT License.

Individual skills declare their license in the license field of SKILL.md (per the Agent Skills spec).

---

Built with ❤️ by the team at Quivly — the AI workforce for post-sales.