
Find strong counterarguments, counterexamples, or failure cases for the current thinking.
- Open source
- 120 GitHub stars

Find strong counterarguments, counterexamples, or failure cases for the current thinking.


Ask the single most important clarification question before answering.

Compare competing ideas or interpretations in the current thinking.

Reduce the current thinking to its smallest useful representation.



Use this skill to generate well-branded interfaces and assets for Honeycomb (shared AI agent memory, a Legion Code × Activeloop product), ei

Use before starting non-trivial work that Honeycomb may already have prior context for (a past decision, a stated convention, where somethin

Honeycomb persistent shared memory for the OpenClaw harness — auto-capture and recall across sessions via the local Honeycomb daemon.

Guide for using the dhub CLI — the AI skill manager for data science agents

Guide for creating effective skills for Claude Code agents

Z.AI CLI providing: - Vision: image/video analysis, OCR, UI-to-code, error diagnosis (GLM-4.6V) - Search: real-time web search with domain/r

Explore code structure, architecture, files, directories, and component relationships using the OpenTrace knowledge graph

Run the full OpenTrace knowledge-graph pipeline over a folder or repo — index, clustering, highlight surfacing, and the markdown / graphml /

Build, package, and verify DesktopManager NuGet, PowerShell module, and CLI/MCP artefacts through the repo-standard PowerForge entrypoints

Operate and validate Windows desktop state through DesktopManager using MCP first and CLI second

Verify requested Windows chat-app workflows through DesktopManager MCP

Inspect Windows desktop applications and verify requested UI workflows with DesktopManager MCP, screenshots, and window or control evidence


Enforces coding best practices: deep-read before modify, LSP-first navigation, TDD red-green-refactor, minimal changes, self-review, verific

Fetch current library/framework documentation via Context7

Systematic debugging: reproduce → hypothesize → verify → fix

Branch completion workflow: verify tests → present 4 options (merge/PR/keep/discard) → execute → cleanup worktree


Generate professional PDF documents with correct CJK (Chinese/Japanese/Korean) rendering

Full plan lifecycle: deep understanding → write plan with TDD checklist → parallel review → Ralph Loop execution

Multi-level research: built-in knowledge → web search → Tavily deep research API


Agent self-learning: promote recurring episodes to rules, capture complex corrections