Agent Skills by Ravindu
Showing 58 vetted skills indexed across 1 GitHub repositories.
agentic-testing
Run disposable Playwright scripts capturing screenshots, video, and DOM snapshots.
document-release
Automate post-release documentation synchronization across docs, READMEs, and API references.
qa
Automate web app QA sessions with Playwright MCP and multi-mode testing.
accessibility-audit
Audit web applications for WCAG 2.1 issues and generate prioritized remediation reports.
impact-analysis
Identify downstream impacts across hooks, scripts, and installed components after kit infrastructure changes.
receiving-code-review
Translate code-review comments into verified code changes with documented rationale.
ship
Automates end-to-end release workflows from feature branch to main.
pre-deploy-security
Execute a pre-deploy security checklist for production deployments in CI/CD pipelines.
pre-landing-review
Automate pre-landing PR reviews with CRITICAL and INFORMATIONAL checks.
security-review
Identify and document security risks in codebases using OWASP Top 10 and threat modeling.
browser-control
Automate browser navigation, clicks, and data extraction via PinchTab commands.
autodev
Automate iterative performance optimization of a single file using benchmark and metric commands.
automation-recommender
Analyze a codebase to identify automation opportunities and generate a prioritized improvement report.
claude-md-improver
Audit CLAUDE.md files against an eight-dimension rubric and output prioritized recommendations.
prompt-engineering
Formalize prompt techniques for structured, reliable AI responses.
using-superpowers
Identify and invoke relevant skills using YAML frontmatter metadata.
ai-agent-workflows
Coordinate autonomous coding agents with HITL checkpoints and safety guardrails.
writing-skills
Author and test Claude Skills using a TDD-inspired RED-GREEN-REFACTOR workflow.
system-architecture
Design scalable multi-service architectures with API decisions, data modeling, and resilience patterns.
cost-aware-llm-pipeline
Route tasks across Opus, Sonnet, Haiku, Ollama Cloud, and Ollama Local to minimize token costs.
dispatching-parallel-agents
Coordinate parallel agent workflows with dependencies and inbox messaging.
brainstorming
Generate structured design concepts from vague ideas with phased guidance and templates.
plan-ceo-review
Plans architecture reviews using EXPAND, HOLD, and REDUCE modes to align technical work with business goals.
content-hash-cache
Cache expensive inputs and model results using SHA-256 keyed entries.