Skill UI

Turns one-shot external recon into a continuous monitoring program: scheduled re-scan-and-diff loop, adversary CTI chatter monitoring across six feeds, finding-lifecycle state machine with SLA and alert dedup, durable alert outbox, and cron recipes. For retainer/MSSP engagements to detect perimeter drift and telegraphed alerts.

Design branching dialogue and narrative graphs with conditions, variables, and localization hooks, and choose between Ink, Yarn Spinner, or a custom data-driven runner. Engine-neutral guidance for wiring conversations into a game loop.

Start or monitor an evolutionary loop that refines ontology and acceptance criteria across generations. Supports starting new loops, monitoring status, rewinding to previous generations, and managing AI agent workflows for iterative software development. It integrates with MCP tools for ontology refinement and acceptance criteria validation.

Build a first-person shooter game with move+mouse-look controllers, hitscan or projectile shooting, weapons, health, and enemy AI. Tune aim feel, time-to-kill, recoil, and spread. Covers core combat loops, weapon systems, and feedback mechanisms for FPS titles.

Reads a session goal, routes it to the right phase, and compiles Claude Managed Agent loops or workflows. Supports goal tracking, bounded iteration, cron scheduling, and BYOK hand-off.

This skill verifies completed work against acceptance criteria using a three-stage pipeline. It performs mechanical checks like linting and testing, semantic reviews for compliance, and multi-model consensus when uncertain. It supports automated fix loops and ensures code quality beyond simple test passing, providing proof of completion.

This workflow implements a bounded, iterative refinement loop for AI agents. It utilizes a required markdown rubric to grade the agent’s output in an isolated context. Based on the grader's verdict, it recommends the next move (sharpen, re-run, or promote). Once the agent passes the rubric, held-back evaluation cases are run in parallel to ensure generalization before final deployment.

This skill manages the human-in-the-loop verification phase of an agent workflow. It ensures that critical artifacts—such as plans, specifications, or migrations—receive mandatory, structured human sign-off before being considered complete or shipped. It collects severity-graded, machine-parseable feedback (e.g., BLOCKER, MAJOR) and acts as a strict gate, preventing completion until all objections are resolved and the review is formally signed off.

A comprehensive guide for setting up and debugging the fundamental structure of a LÖVE (Love2D) game in Lua. Covers the essential callback loop (love.load, love.update, love.draw), delta-time movement, handling various input types, and managing screen states.

Executes an existing Ouroboros Seed spec as a background job, verifies every acceptance criterion, and manages evaluation and Ralph loops.

Configure and build PixiJS v8 render layers by creating async Applications, loading textures via Assets, composing scene graphs with Containers and Sprites, driving ticker loops, handling pointer events, and grouping draws with render groups. Use this skill for initializing PixiJS v8 projects, debugging canvas issues, and managing 2D graphics rendering pipelines.

A comprehensive pattern for developing board-based puzzle games (e.g., Match-3, Sokoban, Logic Grids). This skill focuses on modeling the board state, move input validation, and the core rule resolution loop. It defines how cascades, scoring, and undo functionality must operate deterministically to ensure fair and bug-free gameplay, separating game logic from visual rendering.