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.

A comprehensive utility for managing complex cmux topologies, including workspaces, panes, and surfaces. This tool is essential for inspecting, controlling, and monitoring multiple AI coding agents running in parallel within a macOS terminal environment. It supports advanced functionalities such as browser automation and real-time status notifications, optimizing developer workflows.

This guide details the comprehensive workflow for distributing a single AI skill across various local agent skill directories (Codex, Claude, Pi, Hermes). It addresses canonical locations, symlink management, and specific synchronization commands to ensure that all agents detect the latest skill version while correctly navigating common development pitfalls.

This skill is designed to mitigate false-positive safety triggers from strict AI classifiers (e.g., cyber, bio/chem, reasoning). It operates by making minimal, surgical edits to sensitive phrases, abstracting away specific domains or actions while preserving the original benign intent and structure. It ensures the prompt remains policy-compliant without losing its core purpose, making it ideal for academic or defensive inquiries.

This skill helps developers create highly focused, folder-specific documentation (`CLAUDE.md` and `AGENTS.md`). It systematically analyzes all files in a target directory to extract project purpose, operational constraints, naming conventions, and critical decisions. This deep context ensures that future AI agents (like Claude or Codex) operate with localized knowledge, preventing them from relying on global, generalized instructions and improving their accuracy in that specific domain.

This guide explains how to register custom model slugs (variants) within the Pi Agent's configuration files. It is essential for ensuring that specialized OpenRouter variants or brand-new provider models, which are not part of the static bundled registry, resolve correctly and are available for accurate agent model invocation.

This skill acts as a critical reality-check layer before code generation. It ensures that every function call, import, method, and variable the agent uses is provably available in the user's specific installed environment and version. It prevents runtime bugs caused by hallucinated or improperly assumed APIs, making the generated code robust and reliable.

This skill guides the secure management and operation of Virtual Private Servers (VPS) and deployed server-side agents. Use it when SSH access, system restarts, log inspection, or complex workflow automation (like n8n or custom agents) is required on authorized remote hosts. It emphasizes following strict operational protocols and obtaining explicit user permission for sensitive actions.

This tool reviews code changes (diffs) against existing, accepted Architecture Decision Records (ADRs). It detects potential compliance violations, architectural drift, and non-adherence to established team decisions. Use it before merging Pull Requests or during periodic audits to ensure the code respects agreed-upon architectural constraints and patterns.

Design a persistent multi-agent operating system on Claude Code with kernel routing, specialist agents, slash commands, file-based memory, and scheduled automation for long-lived projects.

This tool automatically generates comprehensive API documentation directly from source code, supporting TypeScript and JavaScript. It scans all public exports (functions, classes, interfaces) and extracts detailed JSDoc annotations. Furthermore, it can generate structured OpenAPI 3.0 specifications for HTTP endpoints, ensuring all public APIs are documented and compliant with industry standards.

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.