claude-skills-marketplace and skill-forge
A provides pre-built Claude skills for specific software engineering tasks, while B is a framework for creating, testing, and publishing custom Claude skills—making them complements where B could be used to develop skills like those in A.
About claude-skills-marketplace
mhattingpete/claude-skills-marketplace
Claude Code Skills for software engineering workflows - Git automation, testing, and code review
Based on the README, here's a technical summary: Provides modular Claude Code plugins (Engineering Workflows, Visual Documentation, Code Operations, Productivity Skills) that activate contextually through declarative SKILL.md and AGENT.md files—enabling Git automation, testing, code review, visual diagram generation, and code refactoring without explicit user invocation. Implements the Anthropic code execution pattern via a sandboxed FastMCP server runtime, achieving 90-99% token reduction for bulk operations by executing Python locally rather than processing files through context.
About skill-forge
AgriciDaniel/skill-forge
Ultimate Claude Code skill creator — design, scaffold, build, review, evolve, and publish production-grade AI agent skills
Implements a multi-tier skill architecture (Tiers 1–4) with specialized sub-skills for planning, building, reviewing, and benchmarking, complemented by dedicated agent roles for design, validation, and execution. Includes cross-platform conversion to OpenAI Codex, Google Gemini, and Cursor, plus integrated evaluation pipelines with variance analysis and multi-agent grading. Built on the open Agent Skills standard with stdlib-only Python implementation for scaffolding, quality auditing (0-100 health scores), and performance benchmarking.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work