Skill UI

The Agent Harness provides a robust, verifiable loop structure for driving complex goals across specialized domains. It converts a high-level objective into a verifiable task plan, executes tasks sequentially using domain-specific tools, and rigorously verifies every step with machine-run checks. The system includes mechanisms for retries, budget management, and mandatory human escalation, ensuring no goal is closed until all progress is verifiably complete or explicitly waived.

Provides deep semantic code intelligence for the Go language, utilizing the official `gopls` language server. Key features include precise go-to-definition, finding references, analyzing call graphs, performing safe refactoring (rename, extract, inline), generating diagnostics, and formatting code. It is essential for navigating, understanding dependencies, and maintaining large Go codebases.

This skill guides the safe, systematic, and at-scale refactoring of existing Go codebases. It ensures that structural changes maintain observable behavior by following a rigorous process: mapping blast radius, strengthening safety nets (tests), applying small tool-driven transforms, and committing changes atomically via staged PRs. Ideal for cleaning up large functions, resolving code smells, renaming packages, or tackling complex architectural improvements in Go.

This skill enforces highly efficient coding practices, guiding the AI to write the simplest, most direct code possible. It strictly follows an 'Efficiency Ladder' (YAGNI, reusing existing code, standard library, etc.) to prevent over-engineering, unnecessary abstractions, and boilerplate. Use this when the goal is maximum efficiency with minimal code.

WGM is a robust, agentic methodology designed to transform vague or rough requirements into working, validated software. It enforces a disciplined workflow across six stages: deep alignment (Grill), structured planning, deterministic iterative building (Loop), mandatory holdout-scenario judging, and final documentation audits. It ensures that the resulting code converges against rigorously tested acceptance criteria rather than relying on single-pass generation.

This comprehensive guide outlines a rigorous, human-in-the-loop engineering workflow for leveraging advanced AI agents to complete complex implementation tasks. It mandates defining clear completion criteria, decomposing work into manageable units, and strategically routing models (Haiku, Sonnet, Opus) based on task complexity. The workflow emphasizes continuous evaluation, regression testing, security review of invariants and edge cases, and strict cost discipline to ensure robust and reliable software development using AI assistance.

This model provides an engineering operational framework for teams that leverage AI agents to generate large volumes of code. It shifts the focus from simple syntax to process quality, robust architecture design, rigorous testing, and systematic risk control, ensuring the development of highly stable and reliable systems in AI-assisted environments.

This function executes a full, autonomous iteration of the Claude Code workflow. It systematically checks the current status, predicts the optimal next action (e.g., editing code, running tests, spawning agents), executes the task, logs the results, and schedules the next wakeup. It discovers and processes tasks from team lists, swarm lists, and file checklists, ensuring continuous and automated progress toward a defined goal.

This skill provides a rigorous, structured method for inspecting animated 3D characters, rigs, and motion retargeting in Blender. It moves beyond visual evidence (screenshots) by extracting and comparing structured data points such as bone transformations, bounding box changes, ground contacts, and axis conventions. Ideal for diagnosing subtle animation flaws like foot sliding, ground penetration, or incorrect orientation that are invisible in still images.

This tool enables web interaction using natural language descriptions instead of complex CSS selectors. By utilizing a specialized page-agent and an LLM-driven tool-call loop, users can simply state their goal (e.g., "Fill the search box with cats and submit"). It is ideal for dynamic web pages, ambiguous structures, or quick, one-shot interactions where robust selector writing is impractical. The system provides graceful degradation if the advanced page-agent is unavailable.

Code Health MCP provides structural maintainability feedback and design-level health scores, going beyond standard linting. It is crucial for AI-assisted coding workflows, allowing users to review code quality, verify score deltas after changes, and gate commits or pull requests to prevent codebase regression. Use it when refactoring, updating legacy modules, or ensuring the integrity of large diffs.

This tool detects cost regressions by comparing two cost-summary JSON snapshots (baseline vs. current). It is essential for PR-gate workflows, checking for unexpected spending increases using three orthogonal signals: total percentage change, absolute dollar increase, and disproportionate growth in specific token classes (e.g., cache writes).