Agent Skills by Svenja Dev
Showing 23 vetted skills indexed across 1 GitHub repositories.
clarify-spec
Generates structured JSON task definitions from vague instructions via clarifying questions.
prompt-architect
Converts vague requirements into structured, best-practice prompts for Claude 4.x.
kaizen
Applies Lean Kaizen methods to analyze and document software problems.
multi-agent-patterns
Distribute tasks across multiple language-model agents using supervisor, swarm, or hierarchical coordination patterns.
preview-testing
Automate end-to-end and security testing for Vercel Preview Deployments with Playwright.
safe-git-guard
Enforce backups and confirmations before destructive Git operations.
multi-llm-advisor
Provides second opinions from Codex 5.1 Pro and Gemini 3 Pro for architecture, code review, and debugging.
ui-freeze
Locks critical UI files and requires explicit confirmation before modifications.
qa-checklist
Automate six-phase QA checks for code changes before merging and deploying.
tdd-strict
Enforce strict test-driven development by blocking code generation until tests exist.
linkedin-engagement
Generate region-specific LinkedIn posts, articles, and comments for manufacturing B2B audiences.
strict-typescript-mode
Enforce explicit types and strict compiler options in TypeScript projects.
gemini-image-gen
Generate branded images from prompts using Google's Gemini API.
e2e-testing
Automate end-to-end browser testing for web applications with Playwright MCP.
changelog-generator
Analyze Git history and convert commit messages into categorized release notes.
project-development
Plan, prototype, and deploy LLM projects using a canonical pipeline.
social-media-content
Generate platform-optimized social media posts for LinkedIn, X/Twitter, Discord, and Reddit.
code-quality-gate
Enforce automated quality checks across pre-commit, CI, preview, E2E, and production stages.
reddit-research
Scan configured subreddits for manufacturing insights and draft replies.
skill-security-analyzer
Analyze Claude skills for security risks and vulnerabilities.
token-budget-advisor
Estimate token consumption and recommend chunked execution plans for large tasks.
learning-capture
Capture reusable patterns and domain knowledge from work sessions into draft skills.
skill-doc-generator
Generate README.md files from SKILL.md metadata with validation.