Executive CPO Agent. 제품 전략, 로드맵, 사용자 가치를 담당합니다.

Executive CTO Agent. 기술 전략, 아키텍처 결정, R&D 리더십을 담당합니다.

>-

exploration-continue

Select mode, delegate to analysis skill.

**Parallelize exploration aggressively.** When researching or exploring, launch multiple subagents simultaneously instead of sequential searches. This reduces wall-clock time proportionally to the num

faber-director-legacy

factory-assembler-expert

Feature store management skill for online/offline feature serving, feature registration, and training-serving consistency.

The `block-commands` skill documents how to use `.claude/settings.json` deny rules to block dangerous commands. However, `.claude/settings.json` is a repository-local file that is not distributed with

Break down features into implementable tasks

feature-documenter

Automate the complete lifecycle of feature flags in Saberloop, following Epic 10 hygiene standards

Count features marked as @failing and write to status JSON file. Used to determine when the autonomous coding loop should end.

**Au démarrage, vérifier l'environnement Figma et les composants existants.**

Find pages with overfull boxes (content that doesn't fit) in LaTeX presentations. Use when checking for layout issues or content overflow.

Fix line endings AND check bash syntax in one step (recommended). Use after creating or editing bash scripts.

Fix CRLF (Windows) to LF (Unix) line endings in files. Use when encountering `/bin/bash: line 1: $'\r': command not found` errors.

Stop asking "How long will this take?" and start asking "How much is this worth to us?" By setting a fixed budget (an appetite) rather than a variable estimate, you force teams to find the simplest, m

Manage Flipt feature flags - list, create, enable/disable, and configure rollout rules. Use when you need to control feature flag state or set up segmented rollouts.

Dart 3.x and Flutter 3.x conventions, naming patterns, code organization, null safety, and async/await best practices

Create Forgekeeper task cards following project conventions

foundations

First-time setup for the foundry plugin (plugin:foundry@claude-foundry)

The **GAM Researcher Agent** automates retrieval and synthesis of context from your Transmission Packet archive. It eliminates manual context pasting by implementing an iterative research loop that se

Analyze large codebases using Gemini CLI's massive context window. Use when analyzing entire codebases, comparing multiple large files, understanding project-wide patterns, verifying feature implementations, or when context limits are exceeded.

Bypass blocked sites via Gemini CLI. Use when WebFetch fails with 403/blocked errors on Reddit, LinkedIn, X, Medium, etc.

>

Minimal implementation of Recursive Language Models (RLM) using Gemini 2.0 Flash and a local Python REPL. Enables processing of massive contexts via the Gemini CLI.

**Purpose:** Strategic enhancement and functional evolution of agent skills

Local storage with GetStorage for preferences, caching, and offline-first patterns

GF(3) Tripartite Orchestration

Creates a new GitHub issue. Use when a new task, bug, or feature request needs to be tracked.

**Goal:** Stage a single, cohesive set of related file changes for an

Create, switch, and manage git branches with proper handling of uncommitted changes. Use when creating feature branches, switching contexts, or cleaning up after merges.

Use when asked to show fetch the context of the current branch

git-safety-guard

**`GOAL`**: present the git status in a visually appealing and

Manage Gitea wiki pages. Use when working with wiki content, creating documentation, or updating wiki pages.

Creates well-structured GitHub issues for the MCPSpy project using the gh CLI tool. Use when asked to create issues, report bugs, or document features. Follows conventional naming with feat/chore/fix prefixes and maintains appropriate detail levels.

GitHub issue lifecycle management with worktree isolation

Spawn and coordinate parallel Claude Code sub-agents using git worktrees. Use this skill when you need to parallelize work across multiple independent tasks, delegate subtasks to sub-agents, or orchestrate complex multi-part implementations.

List and manage cross-session goals and milestones. Use for tracking multi-session work, viewing progress, or managing milestones.

Tracks micro-sprint completion state and progress for Google Sheets operations

GPU-aware PPO training configuration for A100/H100. Trigger when training is slow or GPU utilization is low.

Port of .claude/commands/ground.md for Codex.

Experimentation Agent. A/B 테스트 설계, 가설 검증, 통계 분석을 담당합니다.

Growth Hacker Agent. 성장 전략 수립, 빠른 실험 설계, 그로스 해킹을 담당합니다.

Retention Manager Agent. 사용자 유지, 참여도 향상, 이탈 방지를 담당합니다.

Run diagnostic checks on GSD installation and report status