convex

Q: How to install convex?

Create the skills directory: mkdir -p .claude/skills. Then run: mkdir -p .claude/skills && curl -o .claude/skills/convex.md https://raw.githubusercontent.com/jovermier/claude-code-plugins-ip-labs/main/SKILL.md. Finally, /convex in Claude Code.

Q: What is convex best for?

convex is categorized under Development. It covers: agent.

New

1Community RegistryDevelopmentby Innovative Prospects

Claude Code skills for Convex development in Coder workspaces. Self-hosted Convex, queries, mutations, React integration, RAG, and Chef agent workflows.

First seen 6/3/2026

View Source

Overview

Claude Code marketplace for Innovative Prospects development workflows.

Overview

This marketplace provides plugins, skills, agents, workflows, and commands for full-stack development with comprehensive code review and quality assurance capabilities.

Version: 1.1.0

Key Features:

•13 Specialized Agents (9 review + 4 research) - Parallel agent orchestration
•File-Based Todo System - Persistent work tracking with YAML frontmatter
•Quality Gate Severity Levels - P1/P2/P3 classification for issues
•Knowledge Compounding - Solution documentation for future reference
•Integrated Meta-Workflow - Agents automatically used during development

Plugins

Plugin	Agents	Commands	Skills	Workflows	Description
dev	13	11	8	7	Core workflows, agents, commands, skills, and todo management
coder	-	-	2	-	Coder workspace management and hahomelabs configuration
astro	-	-	1	-	Astro v4-v5 features, islands, content collections
react	-	-	1	-	React 19, React Compiler, modern hooks and patterns
nextjs	-	-	1	-	App Router, Server Components, performance
convex	-	-	4	-	Self-hosted Convex backend development
nhost	-	-	2	-	Hasura, PostgreSQL extensions, and Docker CLI
graphql	-	1	1	-	GraphQL workflow with codegen and type safety
playwright	-	-	2	-	E2E testing patterns and best practices

Totals: 13 agents, 12 commands, 22 skills, 7 workflows

Installation

Add this marketplace to Claude Code:

bash

/plugin marketplace add [your-org]/ip-labs-marketplace

Then install individual plugins:

bash

/plugin install dev@ip-labs-marketplace
/plugin install coder@ip-labs-marketplace
/plugin install astro@ip-labs-marketplace
/plugin install react@ip-labs-marketplace
/plugin install nextjs@ip-labs-marketplace
/plugin install convex@ip-labs-marketplace
/plugin install nhost@ip-labs-marketplace
/plugin install graphql@ip-labs-marketplace
/plugin install playwright@ip-labs-marketplace

To update the marketplace and installed plugins:

bash

/plugin marketplace update ip-labs-marketplace

Quick Start

1. Project Setup

Generate a project-specific CLAUDE.md file:

bash

/project:setup

2. Create a Plan

Start a new feature with a structured plan using parallel research:

bash

/workflows:plan Add user authentication with JWT

3. Execute Work

Implement with quality gates and agent scrutiny:

bash

/workflows:work plans/active/2026-01-12-add-auth.md

4. Review Code

Run comprehensive parallel agent review:

bash

/workflows:review

5. Track Work

Manage todos with the file-based system:

bash

/todo list
/todo create

Dev Plugin Components

Agents (13)

Agent	Purpose
`architecture-strategist`	System design validation and architectural patterns
`agent-native-reviewer`	AI accessibility and agent parity verification
`code-simplicity-reviewer`	YAGNI enforcement and over-engineering detection
`data-integrity-guardian`	Referential integrity and transaction boundaries
`data-migration-expert`	Database migration safety and ID mapping validation
`deployment-verification-agent`	Pre-deployment checklists and rollback procedures
`pattern-recognition-specialist`	Design patterns and anti-patterns analysis
`performance-oracle`	Performance optimization and bottleneck analysis
`security-sentinel`	OWASP compliance and vulnerability scanning

Agent	Purpose
`repo-research-analyst`	Repository structure and pattern analysis
`best-practices-researcher`	External best practices research
`framework-docs-researcher`	Framework documentation lookup
`git-history-analyzer`	Historical context and code evolution

Commands (11)

Command	Purpose
`/workflows:plan`	Create structured project plans with parallel research
`/workflows:review`	Multi-agent parallel code review
`/workflows:compound`	Document solved problems as knowledge
`/workflows:work`	Structured work execution with quality gates
`/deepen-plan`	Enhance existing plans with research
`/todo`	File-based todo management

Command	Purpose
`/project:setup`	Generate project-specific CLAUDE.md
`/project:new-page`	Create new page with component
`/project:new-component`	Create new component with test
`/project:new-blog-post`	Create new blog post
`/test-ui`	Test UI component changes

Skills (8)

Skill	Purpose
`file-todos`	File-based todo system with YAML frontmatter
`quality-severity`	P1/P2/P3 quality gate classification
`plan-manager`	Plan lifecycle management
`context7-skill-generator`	Generate skills from Context7 MCP documentation
`skill-generator`	Generate production-ready Claude Code skills from descriptions
`skill-reviewer`	Review skills for quality, best practices, and PDA compliance
`skill-optimizer`	Refactor skills using Progressive Disclosure Architecture (80-95% token savings)
`skill-architect`	Meta-orchestrator coordinating complete skill lifecycle workflows

Workflows (7)

Workflow	Purpose
`meta-workflow`	Project orchestration with integrated agent scrutiny
`tdd-workflow`	Test-driven development process
`ui-iteration-workflow`	UI component iteration with screenshots
`bug-fix-workflow`	Systematic bug fixing process

Quality Gates

Issues are classified with severity levels:

Severity	Description	Examples
P1 (Critical)	Blocks merge	Security vulnerabilities, data corruption, breaking changes
P2 (Important)	Should fix	Performance issues, architectural concerns, code clarity
P3 (Nice-to-Have)	Enhancement	Code cleanup, optimizations, documentation

File-Based Todo System

Todos are stored as markdown files with YAML frontmatter:

markdown

---
status: pending
priority: p2
issue: "123"
dependencies: []
created: 2026-01-12
updated: 2026-01-12
---

# Add JWT Authentication

## Problem Statement
Application has no authentication.

## Acceptance Criteria
- [ ] Users can log in
- [ ] JWT tokens validated
- [ ] Logout works

Architecture

This marketplace follows the Every marketplace methodology:

•Agents = Specialized review perspectives (how to think)
•Skills = Domain knowledge (what to know)
•Workflows = Development processes (how to work)
•Commands = Slash commands for common tasks
•Hooks = Automation triggers

Directory Structure

code

ip-labs-marketplace/
├── plugins/
│   ├── dev/
│   │   ├── agents/
│   │   │   ├── review/          # 9 review agents
│   │   │   ├── research/        # 4 research agents
│   │   │   ├── design/          # (future expansion)
│   │   │   ├── workflow/        # (future expansion)
│   │   │   └── docs/            # Documentation
│   │   ├── commands/
│   │   │   └── workflows/       # Workflow commands
│   │   ├── skills/
│   │   ├── workflows/
│   │   ├── hooks/
│   │   ├── indexes/            # Plan and context index templates
│   │   └── CHANGELOG.md
│   ├── coder/                  # Coder workspace environment
│   ├── react/                  # React framework skills
│   ├── astro/                  # Astro framework skills
│   ├── nextjs/                 # Next.js framework skills
│   ├── convex/                 # Convex backend skills
│   ├── nhost/                  # Hasura/PostgreSQL skills
│   ├── graphql/                # GraphQL workflow and codegen
│   └── playwright/             # E2E testing skills
├── todos/                      # File-based todos
├── plans/                      # Project plans
└── docs/solutions/             # Knowledge base

Contributing

See CONTRIBUTING.md for contribution guidelines.

License

MIT License - see LICENSE for details.

Acknowledgments

This marketplace is heavily influenced by the Every Inc. compound-engineering-plugin.

Install & Usage

Create the skills directory

mkdir -p .claude/skills

Download the skill file

mkdir -p .claude/skills && curl -o .claude/skills/convex.md https://raw.githubusercontent.com/jovermier/claude-code-plugins-ip-labs/main/SKILL.md

Invoke in Claude Code

/convex

View source on GitHub

agent

Security Audits

LicenseUnknownSourceWarnRepositoryPass

Frequently Asked Questions

What is convex?

Claude Code skills for Convex development in Coder workspaces. Self-hosted Convex, queries, mutations, React integration, RAG, and Chef agent workflows.

How to install convex?

To install convex: create the skills directory (mkdir -p .claude/skills), then run: mkdir -p .claude/skills && curl -o .claude/skills/convex.md https://raw.githubusercontent.com/jovermier/claude-code-plugins-ip-labs/main/SKILL.md. Finally, /convex in Claude Code.

What is convex best for?

convex is a skill categorized under Development. It is designed for: agent. Created by Innovative Prospects.