Claude Skills

The Metal skill analyzes codebases to extract their conceptual architecture as standardized skill, agent, and team definitions. It captures the project's purpose (WHAT) and roles (WHO) without implementation details, abstracting the organizational patterns. Use it for onboarding, bootstrapping agentic systems, or studying a project's DNA for inspiration.

The `meditate` skill provides meta-cognitive techniques for developers to clear context noise and improve AI task focus. It helps observe reasoning patterns, manage distractions like scope-creep, and reset attention between unrelated tasks. Use it when your reasoning feels scattered, before deep work, or when assumptions may be biasing your analysis.

This skill guides users through a complete, structured meditation session, from posture selection to session integration. It coaches on techniques like breath anchoring, managing distractions, shamatha, and vipassana. Developers should use it to add meditation guidance for stress management, focus preparation, grounding, or integrating intense experiences.

This skill provides token budget management for long-running AI agent workflows, helping to control costs and prevent context window overflows. It tracks usage, enforces caps, and can prune context or use progressive disclosure when limits are approached. Use it to add cost guardrails to autonomous agent loops, polling systems, or any workflow where context accumulation is a risk.

This Claude Skill helps developers build tools to manage TCG collections by providing inventory tracking, storage guidance, and grade-based valuation for cards. It's useful for cataloging collections, assessing value for insurance or sale, and deciding on professional grading submissions. The skill supports popular games like Pokémon, MTG, Flesh and Blood, and Kayou.

This skill helps developers manage R package dependencies using renv to create reproducible environments. It handles initialization, snapshot/restore workflows, troubleshooting, and CI/CD integration for R projects. Use it when setting up new projects, restoring environments on different machines, or resolving dependency issues.

This skill maintains Claude Code's persistent memory by organizing MEMORY.md as a concise index and extracting detailed topics into separate files. It automatically detects stale entries, verifies accuracy against the current project, and enforces a 200-line limit. Use it when MEMORY.md approaches the line limit, after productive sessions, or when project changes may have outdated memory entries.

This skill provides production-grade Kubernetes secrets management using SealedSecrets for GitOps workflows and External Secrets Operator for cloud integrations. It handles encryption, RBAC, and rotation strategies for TLS certificates, API keys, and credentials. Use it when you need version-controlled secrets, integration with AWS Secrets Manager/Azure Key Vault, or to migrate plaintext secrets to encrypted formats.

This Claude Skill helps developers manage Git branches for features, fixes, and cleanup. It handles creating, switching, syncing with upstream, and pruning merged branches using safe practices like stashing. Use it when starting new work, switching tasks, or keeping branches current with main.

This skill manages a queue for incoming engagement items (like notifications or messages) by ingesting, deduping, prioritizing, and rate-limiting them. It generates periodic digests for processing and enforces cooldown periods, acting as a buffer between platform signals and action cycles. It's designed to compose with a separate timing skill (du-dum), which handles the observation/action cadence.

This Claude Skill helps developers maintain a project changelog following the Keep a Changelog standard. It manages entry categorization (Added, Changed, Fixed, etc.) and handles version sections, including promoting Unreleased entries to versioned sections during releases. Use it when starting a new project, adding entries after changes, preparing releases, or converting existing changelogs to the standard format.

This Claude Skill manages the complete change control process for validated computerized systems in regulated environments. It handles request triage, impact assessment, approval workflows, and post-change verification for software updates, patches, or configuration changes. Use it when you need to maintain compliance during system modifications, including emergency changes requiring expedited approval.

This skill helps developers manage BibTeX bibliographies through R, enabling parsing, merging with deduplication, and generating entries from identifiers like DOI or ISBN. It's useful for creating clean .bib files for R Markdown/Quarto or consolidating bibliographies from multiple collaborators. Key capabilities include intelligent deduplication by DOI/title similarity and exporting sorted, structured BibTeX output.

This Claude Skill helps developers create and maintain a prioritized product backlog using techniques like user story writing and MoSCoW prioritization. It assists in converting project scope into actionable items, grooming the backlog before sprints, and re-prioritizing after feedback. Key features include item splitting, status tracking, and managing acceptance criteria with estimates.

This skill enables starting and maintaining a fire using friction, spark, or solar methods, handling site selection, material grading, and fire laying techniques. It's designed for use cases like providing warmth, light, signaling, boiling water, cooking, and emergency survival. Developers can integrate it for scenarios requiring wilderness survival or primitive fire-making capabilities.

This skill provides maintenance instructions for eight essential garden hand tools, including sharpening, cleaning, and seasonal care routines. It triggers for specific maintenance schedules like after each use, monthly during the growing season, or for seasonal storage. Developers can integrate it to offer structured, actionable tool-care guidance within gardening applications.

The "listen" skill enables Claude to practice deep receptive attention, parsing multiple layers of communication—literal, emotional, contextual, and meta—to extract true intent beyond surface words. It's designed for ambiguous requests, when context contradicts literal meaning, or before major tasks to prevent costly misunderstandings. Developers should invoke it when previous responses have missed the mark or when the full signal requires holistic integration.

The `listen-guidance` skill coaches users in active listening techniques, including fostering a receptive mindset, practicing reflective paraphrasing, and using clarifying questions. It is designed for developers to integrate when users need to improve communication, prepare for difficult conversations, or address frequent misunderstandings. The skill focuses on guiding the practice and integration of these methods into daily communication.

The `learn` skill enables Claude to systematically acquire knowledge from unfamiliar domains using a structured reasoning process of surveying, hypothesizing, probing, and verifying. It is designed for situations like exploring an unfamiliar codebase, investigating topics beyond simple recall, or resolving conflicting information into a coherent model. Key capabilities include deliberate model-building with feedback loops and the use of tools like Read, Grep, and WebSearch for exploration.

This Claude Skill acts as an AI learning coach to guide developers through structured learning of new technologies or skills. It assesses existing knowledge, designs a personalized learning path, walks through material, and tests understanding with adaptive difficulty. Use it when starting a new tech, feeling overwhelmed by documentation, needing spaced repetition for retention, or transitioning domains requiring a gap analysis.

This skill sets up systematic data labeling workflows using tools like Label Studio, including quality control and inter-annotator agreement. It helps integrate labeled data into ML pipelines for text, image, audio, or video when starting supervised ML or when model performance is limited by insufficient labels. It also supports active learning to prioritize valuable examples for labeling.

This skill guides developers through structured root cause analysis and CAPA management for compliance deviations. It provides method selection (5-Why, fishbone), action design, and effectiveness verification. Use it when addressing audit findings, system deviations, or recurring issues requiring a systematic investigation.

This skill boosts Claude's intrinsic motivation to shift from routine compliance to genuine engagement, especially for complex or long-running tasks. It applies psychological frameworks like Self-Determination Theory and Flow to enhance creativity and sustained focus. Use it when work feels formulaic, before a creative challenge, or to rekindle enthusiasm for a project.

This skill analyzes UV-Vis spectra to identify chromophores and classify electronic transitions, applying Woodward-Fieser rules for conjugated systems. It also performs quantitative concentration analysis using the Beer-Lambert law. Use it for confirming structural features like aromatic rings or for monitoring reaction kinetics via absorbance changes.