Improve an existing SKILL.md with forge eval until it passes the quality gate, fixing the weakest category each round.
Run SkillTool slash commands inside a forked sub-agent so the skill has isolated budget, telemetry, and lifecycle.
Use when we want to turn a just-finished Formax workflow (e.g. commands, overlays, tools, hooks, permissions, UI parity) into a reusable Codex Skill under .codex/skills, including…
Liste und Snapshot-Abruf von Webcams (insb. foto-webcam.eu). Verwende diese Skill-Anleitung, wenn John „webcam “ schreibt, eine Webcam-Favoritenliste pflegen will — from…
Before declaring a new or substantively-revised governance document complete, dispatch a fresh subagent with no session context to read the document and surface tacit-context gaps…
Compress an agent's routing file (RESOLVER.md or AGENTS.md) by converting granular skill-per-row tables into functional-area dispatchers.
Use for approved implementation with Superpowers worktrees, TDD, execution plans, subagents, and verification.
The suite scanner, three lenses — codebase health (god files, churn hotspots, coupling, bug concentration, scope creep), decision-debt (absorbed from gabe-debt: decisions never…
Coordinate multiple game-development skills and produce outcome-first briefs, plans, reviews, and validation flows.
Use when running a GPU Category agent end-to-end over the live web — fans out gatherer subagents that follow the trail of leads, snapshots raw documents, then runs the frozen…
Use when a task needs an exceptional/impressive result and you can afford many tokens and time — large creative builds (games, apps, sites, 3D scenes) with subagents.
Analyzes worker pain points to identify missing skills and create skill proposals. Use during playtest GDD review phase when workers report repeated struggles with specific…
Pipeline 5 subagentów analizujących tickery — Scout → Quant → Alpha → Auditor → Director — plus Weekly Tracker (Agent 06) do cotygodniowego audytu skuteczności rekomendacji.
Analyzes agent skills to understand their business purpose and proposes functional enhancements. Identifies opportunities for new features, workflow improvements, better…
Use during the initial-creation iteration loop of -requirements.md / -tech-design.md / -implementation-plan.md.
Use when needing workflow artifacts from PR - runs in subagent to find jobs with artifacts, returns gh commands to download them without consuming context
Audit a drafted or filed GitHub issue / PR body against gh-body-conventions via a fresh-context subagent. Any unresolved ⚠ blocks the caller.
Find and answer unanswered GitHub discussions. Activate when user wants to contribute to open source discussions, help answer GitHub questions, or find discussions to answer.
Turn this Claude Code session into a looping timelapse GIF of the Roundtable pixel office — every subagent as a person who walks in, works, reports and leaves.
Make a GitHub repo or Agent Skill discoverable across the skill-aggregator ecosystem — GitHub topics, scraper-visible SKILL.md paths, instant directories (awesomeskills.dev /…
Manage skills published as GitHub Gists — create, update, auto-update, and handle username changes
Gerencia o ciclo de vida de Agent Skills usando os comandos oficiais 'gh skill' da GitHub CLI. Valida, inspeciona e publica pacotes de habilidades diretamente em repositórios do…
Goal-driven harness loop — keep working on a goal until /es, /er, /code_standards, and Codex plugin review all pass. All gates use adversarial subagents. Alias: /h
Use for executing and enforcing governance tooling (artifact generation/validation/CI policy checks), including startup declaration checks, project-index sync checks, and…
Route GPT-5.6 tiers, reasoning effort, and subagents to minimize accepted-result cost subject to an explicit quality floor.
Launch a GPT 5.6 Sol Max subagent to do a deep, neutral senior-developer review of the current work and report its findings back verbatim.
Context-efficient Git and Graphite workflows for Claude Code. Automatically delegates verbose git/Graphite CLI operations to isolated subagents, reducing context pollution by…
Comprehensive implementation auditor that grills code quality, plugin structure, consistency, and identifies problems.
Ordnet unübersichtliche WEG- und Hausverwaltungsakten mit vielen Konflikten: Heizung, Dach, Gewerbe, Geruch, Tauben, Fahrrad, Kinder, Wallbox, Steckersolar, Nebenkosten,…
Use when measuring a cc-master skill — 当你写完一个 skill 的 body 要度量它好不好、要迭代它、或在改 description / 纪律段前后想知道改动有没有用时.
Traces decisions and requirements through GSD artifacts. Use when user asks "why did we...", "what happened to...", or mentions decision history.
Interrupt controller converting polled to proactive agent execution. Per-runtime strategies, configurable thresholds, Deacon heartbeat supervision.
Han projesinin yazım, naming, commit, error handling ve test kuralları. Tüm subagent'lar bu kurallara uyar.
針對單一 harness-eval 維度做深度稽核與具體修法。配合 /harness-eval 使用： 先跑全面評估，發現 WARN/FAIL 後用此 skill 精準挖掘。 用法：/harness-eval-focus D2（或 D1~D10）。 觸發關鍵字：harness-eval-focus、深度稽核、維度修法、D2 hook 問題、 D3…
Improve Kandev's AI harness from session learnings or explicit requests. Use when the user asks to record learnings, update or create skills, agents, subagents, commands,…
Comprehensive multi-stage code review using specialized subagents
Representations for hierarchical skill structures including knowledge graphs and ontological decomposition
Use for affordances, hierarchical control, nullspace composition, and information-seeking action selection in embodied agents.
Intelligent task routing system that identifies task types (web/code/data/doc/chat) and routes to optimal models with appropriate execution mode (subagent/main session).
Audit hook delivery health from Agent Monitor data — balance PreToolUse vs PostToolUse (a gap means tools that started but never reported back), detect missing Stop/SubagentStop…
Run ComfyUI workflows from any AI agent (Claude Code, OpenClaw, Codex, Hermes) via a single CLI. Import workflows, manage dependencies, execute across multiple servers, a — from…
Orchestrator-driven parallel implementation — plans tasks, dispatches worker subagents, verifies, commits atomically per task.
Execute a plan task-by-task using fresh subagents, with spec compliance + code quality reviews between tasks.
Stress-tests skills with edge-case prompts and scoring rubric. Use for: test skill, hyper test, skill QA, audit skill quality.
Use when an agent violates a loaded skill despite it being present in context — a gap exists between what the skill says and what the agent did.
Use when analyzing Claude Code conversation logs to find patterns in repeated user instructions that could become skills. Ask for date range first.
GLM-4.6v image analysis subagent for game screenshot and sprite quality assessment
Use when executing implementation plans with independent tasks. Risk-aware subagent dispatch with per-task review.
Shared rules for all implement subagents in REDACTED-webserver. Layer order, file scope limits, verify commands, and skill loading.
Single-agent implementation: executes plans in-place without subagents. Use /implement (default) for complex features; use this for simpler tasks or when subagent overhead is…
Drain the ready-issue backlog with quality: claim a batch of independent issues, implement each in parallel TDD worktree subagents, then serially merge green PRs.
Implement one planned unit of work in a subagent, then commit, merge, and push it with error recovery
Orchestrates parallel implementation of ready-for-agent frontier tickets with isolated workers and orchestrator-owned review and commit. User-invoked via /implement-with-subagents.
Execute implementation plans by delegating each task group to task-group-implementer subagent. Main agent coordinates prepares context, invokes subagent, processes output, marks…
Use when implementing skill data generation from HTML sources for game build planners - guides the parser-factory-generation pattern for extracting level-scaling values for active…
Build the standardized prompt for a fresh reviewer subagent that evaluates implemented work against its spec without access to the implementation conversation.
Iterative benchmarking, evaluation, and algorithmic optimization of alternative A2UI inference formats (such as Express, Atom, and Elemental).
Export and validate InferMind skill-driven run artifacts, including InferIR, trajectory summaries, stage decisions, and validation reports.
Use when reviewing or editing proposed or existing public skills for structure, metadata, source safety, usefulness, validation, and release readiness.
Mark first-time skill listings differently from later deltas so attachment consumers can distinguish bootstrap exposure from incremental updates.