Agent Skills by gwozai
Showing 20 vetted skills indexed across 1 GitHub repositories.
gwozaitroubleshoot
Guide users through root cause analysis for software errors and bugs.
gwozaitoken-efficiency
Compresses AI-generated output tokens to maintain quality in low-context sessions.
gwozaitemplate-skill
Create structured task templates for Claude Code workflows using SKILL.md.
gwozaidoc-coauthoring
Guide structured documentation co-authoring through context gathering, refinement, and reader testing.
gwozaiclaude-api
Build and optimize Claude API applications with automated prompt caching.
gwozaixlsx
Create and edit spreadsheet files using Python's pandas library.
gwozaiExtract text and structured table data from PDF documents using pypdf and pdfplumber.
gwozaiinternal-comms
Generate internal communications documents from predefined templates and formatting guidelines.
gwozaiskill-creator
Guides users through creating, modifying, and optimizing AI skills.
gwozaicanvas-design
Generate visual art concepts guided by a design philosophy.
gwozaipptx
Create and edit PowerPoint presentations using PptxGenJS and LibreOffice.
gwozaifrontend-design
Develop distinctive frontend interfaces with bold aesthetic direction and technical implementation.
gwozaimcp-builder
Develop and configure MCP servers in Python or Node/TypeScript.
gwozaidocx
Create, edit, and analyze .docx files with formatting and content extraction.
gwozaiusing-git-worktrees
Creates isolated Git worktrees for feature branch development with safety checks.
gwozaiexecuting-plans
Execute pre-written implementation plans with review checkpoints and error handling.
gwozaifinishing-a-development-branch
Guides users through verifying tests, merging, creating PRs, and cleaning up git branches.
gwozaibrainstorming
Guide structured brainstorming to develop detailed design specifications.
gwozaiwriting-skills
Create skill documentation using test-driven development principles.
gwozaisubagent-driven-development
Execute development plans by dispatching subagents for individual tasks with two-stage reviews.