Agent Skills Hub

create-unit-test

Create and run unit tests for AnySoftKeyboard following project architecture guidelines (Robolectric, naming, location).

qa-tester

Specialized QA testing agent for morphir-dotnet, covering test plans, regression, E2E verification, bug reporting, and package validation.

find-skills

Discover and install agent skills to extend your DeerFlow capabilities. Use this to find tools, workflows, or specialized knowledge for tasks like coding, testing, and deployment.

epic

Create and manage Epics for organizing large-scale feature development, themes, and related issues into a structured project hierarchy.

analyzing-research-documents

Systematically extract insights, decisions, and constraints from research documents, technical papers, and architectural design files.

deep-research

A systematic, multi-angle web research agent. Use for deep investigation, complex queries, and as a mandatory pre-research step before content generation to ensure evidence-backed, high-quality results.

git-worktrees

Manage Git worktrees for parallel feature development. Automates environment setup, branch hierarchy enforcement, and workspace cleanup based on Langstar's issue-driven workflow.

evidence-first-debugging

Enforces a strict evidence-based debugging workflow using structured observation, hypothesis testing, and causality validation to eliminate speculation in technical investigations.

swiftui-performance-audit

Audit and optimize SwiftUI runtime performance. Diagnose slow rendering, janky scrolling, high resource usage, and layout thrash with code review, architecture analysis, and Instruments profiling guidance.

find-skills-combo

Discover and recommend combinations of agent skills to complete complex, multi-faceted tasks using Maximum Quality or Minimum Dependencies strategies.

oh-my-claude

Activates Prometheus planning mode for structured requirement gathering, codebase research, and task planning within Claude Code.

hs

Pre-execution security guardrails for AI agents. Validates shell commands and file reads against 400+ security patterns to block destructive operations, credential theft, and unauthorized system access.