Testing skills

This skill audits skill.md style instructions for supply-chain risks, exposing exfiltration patterns and generating a safe markdown report with risk level.

This skill generates a concise codebase overview including language, framework, architecture, entry points, and key files for onboarding.

This skill enforces test-driven loops for non-deterministic agents, ensuring tests run before production code and blocking warnings as errors.

This skill automates Linux X11 desktops for GUI tasks like capturing screens, locating elements, clicking, typing, and window management.

This skill guides a complete cybersecurity assessment and threat modeling to harden systems across any stack, with actionable mitigations.

This skill helps automate ML workflows by guiding task runners, containerization, CI/CD, and experiment tracking across projects.

This skill helps you manage login state and switch accounts for Xiaohongshu by controlling CLI flow and cookies.

This skill helps you develop and extend Lobster AI by guiding planning, scaffolding, and contributing to agents, services, and plugins.

This skill provides an eye to browse the web with a real Chromium browser, rendering dynamic content and taking screenshots for verification.

This skill helps you evolve code with a self-improving agent, leveraging multi-LLM guidance to generate, test, and crystallize reusable capabilities.

This skill helps determine semantic version bumps (major, minor, patch) and aligns releases with SemVer 2.0.0 guidelines.

This skill performs a deep security audit with 1,000 iterations to uncover critical risks and strengthen pre-release compliance.

This skill analyzes JUnit retry artifacts to identify flaky tests and produce a prioritized triage report.

This skill helps you quickly test and switch MCP preview modes, validate iframe rendering, and trigger external HTML delivery for Claude MCP tools.

This skill benchmarks your agent memory retrieval using a peer-review-grade suite, collects anonymized statistics, and submits results to research aggregations.

This skill guides your team through a TDD-first, pipeline-driven task workflow with planning approvals, Trello-style tracking, and PR-based reviews.

This skill clusters JUnit failures into stable fingerprints to accelerate root-cause triage by ignoring noisy logs.

This skill launches two Chrome instances, one normal and one with remote debugging on port 9222, enabling easy debugging workflow.

This skill automates cross-platform PSD text updates with safe dry-run, rollback, and local index caching to streamline design workflows.

This skill simulates neural spike activity and decodes it into 2D cursor velocity for Brain-Computer Interface experiments.

This skill tests Rust, Go, and Java projects by auto-detecting type, running tests when present, compiling, and reporting results for fixes.