Claude Code Hooks sind leistungsstark, aber unsichere Hook-Skripte können Daten preisgeben oder Arbeit unerwartet blockieren. Diese Skill erklärt Hook-Events, Konfigurationsmuster und sicherere Skriptvorlagen für kontrollierte Automatisierung.

Teams benötigen wiederholbare Claude Code-Läufe ohne interaktives Terminal. Dieser Skill erklärt die Headless-CLI-Nutzung, strukturierte Ausgabe, CI-Muster und Tool-Steuerungen.

Clerk management tasks require correct authentication, endpoint discovery, and instance targeting. This skill guides CLI workflows for resource management, configuration, webhook testing, and deployment verification.

Astrale operations require precise paths, identity selection, and command options. This reference guides assistants through Kernel operations, graph workflows, diagnostics, and local UI setup.

Langlaufende Entwicklungsserver und Watcher können Ausgaben verlieren oder die Arbeit von Agenten blockieren. Diese Skill bietet tmux-Muster zum Starten, Überwachen, Neustarten und Isolieren dieser Prozesse.

Lokale Tilt-Umgebungen können fehlerhafte Ressourcen, veraltete Logs und Docker Compose-Health-Probleme verbergen. Dieser Skill führt Claude, Codex und Claude Code durch Statusprüfungen, Logs und Tiltfile-Korrekturen.

Parallele lokale Projektkopien können bei Ports, Daten, Cookies und Docker-Ressourcen kollidieren. Diese Skill-Anleitung unterstützt die Erstellung oder Überprüfung eines gen-env-Befehls, der jede Instanz isoliert.

Bei der manuellen PR-Einrichtung können geänderte Dateien, Tests oder Repository-Konventionen übersehen werden. Diese Skill führt durch einen japanischsprachigen Workflow für Review, Commit, Push und Erstellung eines GitHub-PR.

Moving an existing agent workflow into Factory can create duplicate roles, broken paths, or conflicting handoffs. This skill guides a focused migration into an existing Factory repository and validates the resulting configuration.

CI failures can hide the first useful error among checks, logs, and repeated noise. This skill gathers relevant GitHub evidence and produces a prioritized plan for repair.

Per-workspace environment recipes can fail because CLI selection and provider lifecycle details vary. This skill loads Orca guidance matched to the active version.

Local dashboards need a simple signal that a tool is available. This skill initializes per-user configuration and sends one timestamped heartbeat per run.

Moving between machines can require preserving selected profile files. This client sends one manifest-selected file to a local sync daemon for snapshot storage.

Tracking shell profile loads requires a persistent startup marker. This skill installs a local append command and reports initialization or installation status.

Repeated shell setup can create duplicate startup entries. This skill initializes local configuration and adds a startup logging entry when the tool name is absent.

A page needs fetchability and listing evidence before its citation potential can be assessed. This skill organizes a single-URL checklist covering status, robots, sitemap, and scoring validation.

AI client setup often fails because tools expect different base URL and protocol formats. This skill selects the correct RouteMux endpoint and verifies access.

Teams need a faster GitHub Actions runner setup without manually editing every workflow. This skill runs the Peak workflow, guides authentication, and reports the proposed changes for review.

Debugging related microservices requires coordinating local processes, cluster targets, and traffic filters. This skill guides shared YAML configuration and concurrent mirrord session management.

Temporal workers need precise task routing during local debugging, but manual Kubernetes configuration is easy to misconfigure. This skill generates resource guidance, filters, validation checks, and troubleshooting steps for mirrord Temporal splitting.