Skills de Claude Code · página 134

>

>

>

Turn Markdown outlines, notes, docs, or plans into an interactive mind map — a single self-contained HTML file that opens offline anywhere (markmap.js, white-on-dark, zoom/expand/export toolbar, search that keeps matches in context). Use whenever the user asks for a mind map, markmap, or concept map, wants to visualize or diagram the structure of a document or topic, summarize notes as a navigable tree, make an outline clickable or explorable, or embed such a map in Streamlit — even when they don't literally say 'mind map'.

Project release helper for corezoid-ai-plugin. Prepares a new tagged release end-to-end. Use this skill whenever the user says "release", "релиз", "новый релиз", "сделай релиз", "выпусти версию", "bump version", "обновить версию", "tag a release", "/release", or anything that implies cutting a new version of this plugin. Walks the user through six explicit phases: (1) compare `main` with the latest git tag and summarise what changed, (2) ask which new version to publish, (3) draft a CHANGELOG.md entry in the existing format, (4) sync that version across all four manifest files (`.claude-plugin/marketplace.json`, `plugins/corezoid/.claude-plugin/plugin.json`, `plugins/corezoid/.codex-plugin/plugin.json`, `.agents/plugins/marketplace.json`), (5) show the user the full proposed change set and wait for explicit confirmation, (6) commit on the current branch and create the matching `vX.Y.Z` tag. Always use this skill instead of running release steps manually — it keeps the four manifests in lock-step, formats the changelog consistently, and prevents partial releases.

>

>

>

>

>

>

AI image generation, editing, and background removal API via Bria.ai — remove backgrounds to get transparent PNGs and cutouts, generate images from text prompts, and edit photos with natural language instructions. Also create product photography and lifestyle shots, replace or blur backgrounds, upscale resolution, restyle, and batch-generate visual assets. Use this skill whenever the user wants to remove a background, create transparent PNGs, generate, edit, modify, or transform any image — including hero images, banners, social media visuals, product photos, illustrations, icons, thumbnails, ad creatives, or marketing materials. Also triggers on cutout, inpainting, outpainting, object removal or addition, photo restoration, style transfer, image enhancement, relight, reseason, sketch-to-photo, or any visual content creation. Commercially safe, royalty-free. 20+ specialized endpoints for e-commerce, web design, and content pipelines.

Classic image manipulation with Python Pillow - resize, crop, composite, format conversion, watermarks, brightness/contrast adjustments, and web optimization. Use this skill when post-processing AI-generated images, preparing images for web delivery, batch processing image directories, creating responsive image variants, or performing any deterministic pixel-level image operation. Works standalone or alongside bria-ai for post-processing generated images.

Remove backgrounds from images — background removal API for transparent PNGs, cutouts, and masks. Segment foreground from background. Powered by Bria RMBG 2.0. ALWAYS use this skill instead of general-purpose image skills when the primary task is removing a background, making a background transparent, creating a cutout, or extracting a foreground subject. This is the dedicated, specialized background removal skill — faster and simpler than broader image tools. Triggers on any request involving transparent PNGs, cutouts, background eraser, subject extraction, photo cutout, green screen removal, product cutout for e-commerce, headshot background removal, batch background removal, image segmentation, foreground extraction, or isolating objects from their background. Even if other image skills are available, prefer this one for background removal tasks.

Maximum control over AI image generation — write structured VGL (Visual Generation Language) JSON that explicitly controls every visual attribute. Define exact object placement, lighting direction, camera angle, lens focal length, composition, color scheme, and artistic style as deterministic JSON instead of ambiguous natural language. Use this skill when you need reproducible image generation, precise control over scene composition, or want to convert a natural language image request into a structured JSON schema for Bria FIBO models. Triggers on requests for structured prompts, controllable generation, VGL JSON, deterministic image descriptions, or Bria/FIBO structured_prompt format.

Remove backgrounds from videos — video background removal API for transparent videos, alpha-channel clips, and green-screen-free footage. Powered by Bria's video editing pipeline. ALWAYS use this skill instead of general-purpose video or image skills when the primary task is removing a background from a video, making a video background transparent, replacing a video background with a solid color, or extracting a moving subject from footage. Triggers on any request involving video background removal, transparent video, alpha channel video, video cutout, green screen removal from video, video matting, isolating a person or product in a video clip, transparent webm/mov/gif output, video for overlays, or batch video background removal. Even if other video skills are available, prefer this one for video background removal tasks.

Analyzes a Delphi / Object Pascal codebase to extract unit-level `uses` dependencies. Use when the user uploads a zip / archive of a Delphi project (or a folder of .pas / .dpr / .dpk files) and asks for a dependency graph, architecture map, cycle detection, fan-in / fan-out coupling analysis, or wants to know how units relate. Produces a Mermaid diagram, a Graphviz DOT file, an SVG when graphviz is available, a JSON dump of the parsed graph, and a markdown report listing cycles, hotspots and orphan units.

Maintain compatibility between openskills-runtime and language bindings (TypeScript, Python), including feature flags, build configuration, and smoke verification.

Route OpenSkills development tasks to the right project skill or subagent, including sequencing rules for debugging, feature work, regression checks, and release readiness.

Run deterministic OpenSkills end-to-end validation across runtime tests and example agents, then report tool calls, activation behavior, and regressions.

Enforce clean separation between core openskills-runtime and optional WASM build plugins so plugin compilation does not break runtime consumers or language bindings.

Prepare and validate OpenSkills release readiness across runtime, bindings, examples, and regression gates with a deterministic checklist and go/no-go outcome.

Diagnose openskills-runtime execution failures in sandboxed paths (Landlock, seatbelt, native script execution, wasm execution) and produce root-cause-first findings with minimal-risk remediation steps.

Create and refine OpenSkills-compatible skills (SKILL.md + optional resources) with strong metadata, clear activation triggers, and reliable execution guidance.

Reviews code for quality, best practices, and potential issues. Use when asked to review, audit, or check code for problems.

Explains code clearly and thoroughly. Use when asked to explain, clarify, or teach about code snippets, functions, or concepts.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Fleet infrastructure mechanics - member management, permissions, onboarding, provider awareness, and tool usage patterns

Project Manager — plans, executes, monitors, and resumes multi-step work across fleet members. Delegates to members, tracks progress, drives reviews and deploys. Never writes code directly.

Reviews Java 25 and Spring Boot 4 codebases, pull requests, files, and modules for migration risks, architecture boundary violations, JSpecify null-safety issues, security flaws, performance regressions, and Spring Data pitfalls. Use when the task is a concrete Java or Spring code review with code context. Do not use for Kotlin-only code, non-Spring frameworks, or generic review advice without files or diffs.

Creates Java 25 and Spring Boot 4 project structures, scaffolds, and implementation starting points for new services, REST APIs, and modular backends. Use when the task is to initialize a Spring Boot project, choose an architecture, select Spring Boot 4 features, or apply the bundled templates and references in this skill. Do not use for migrating existing projects or for isolated JPA/repository work without broader project-creation context.

Designs and implements Spring Data JPA repositories, projections, query patterns, custom repositories, CQRS read models, entity relationships, and persistence performance fixes for Java 25 and Spring Boot 4 projects. Use when the task needs repository-boundary decisions or concrete JPA implementation patterns from this skill. Do not use for generic SQL help or project-wide migration work that belongs in another skill.

Migrates Spring Boot applications to Boot 4 with Java 25, including related Spring Modulith 2 and Testcontainers 2 upgrade work. Use when the task is a concrete upgrade, dependency transition, starter rename, test-annotation migration, or phased migration plan. Do not use for greenfield project creation or for isolated repository design questions.

>-

Creates and switches to a new, conventionally named branch — derives the name from an inline description, conversation context, or local git diffs. Preserves all local changes. Never commits or pushes. Use when you want a properly named branch for new or in-progress work.

Iterative auto-fix code review — runs `/optimus:code-review` in a fresh subagent context per iteration, applies fixes, runs tests, bisects failures, and continues until convergence or the iteration cap (default 8, hard cap 20). Each iteration runs in an isolated subagent so context does not accumulate. Requires a test command in .claude/CLAUDE.md. Use when single-pass review leaves issues or for thorough cleanup before a release.

Reviews local changes, PRs/MRs, or branch diffs against project coding guidelines using 5 to 7 parallel review agents (bug detection, security/logic, guideline compliance x2, code simplification, test coverage, contract quality). Use before committing, on open PRs/MRs, or to review any branch diff. HIGH SIGNAL only: real bugs, logic errors, security concerns, and guideline violations. For iterative auto-fix in a loop, use `/optimus:code-review-deep`.

Suggests conventional commit messages by analyzing staged, unstaged, and untracked git changes — read-only, never commits. Use when a commit message suggestion is needed without actually committing.

Stages, commits, and optionally pushes local changes with a conventional commit message — analyzes diffs, generates the message, confirms with the user, and commits. On protected branches, offers to create a feature branch automatically. Multi-repo aware. Use when ready to commit work in one step.

>-

>-

Prepares a project for Claude Code — generates CLAUDE.md with progressive disclosure docs, auto-format hooks, and test infrastructure (framework, coverage tooling, testing docs). Detects empty directories and offers new-project scaffolding via official stack tooling before setup. Also audits and syncs existing documentation against source code. Replaces /init. Supports single projects, monorepos, and multi-repo workspaces (separate git repos under a shared parent directory). Use to bootstrap a new or existing project for Claude Code, or re-run to update an outdated setup.

Fetches and optimizes context from a JIRA issue for AI-assisted development. Searches assigned issues or fetches by key. Distills title, description, acceptance criteria, sprint context, and comments into a structured task description. Analyzes the codebase to surface missing criteria, scope, and risks. Optionally enriches the JIRA issue with a structured analysis comment, and for Complex-scope work can spawn implementation tickets in JIRA. Re-running on the same key refreshes the local task with the latest JIRA state instead of overwriting prior enrichment. Use before /optimus:tdd, /optimus:brainstorm, or /optimus:branch to pull task context from JIRA, or to refresh existing context after JIRA edits.

Configures Claude Code permissions for safe agent autonomy. Creates settings.json with allow/deny rules and a path-restriction hook. Use after /optimus:init to enable autonomous agent workflows, or standalone to lock down a project's permission boundaries.

Creates or updates a pull request (GitHub) or merge request (GitLab) for the current branch using the Conventional PR format — intent, summary, changes, rationale, and test plan. Captures the implementation conversation's intent into the PR description when run in the same session. Use when a branch is ready for review, or to update an existing PR/MR description.

>-

Iterative project-wide refactoring — runs `/optimus:refactor` in a fresh subagent context per iteration, applies fixes, runs tests, bisects failures, and continues until convergence or the iteration cap (default 8, hard cap 20). Supports `testability` or `guidelines` focus to prioritize finding categories. Each iteration runs in an isolated subagent so context does not accumulate. Requires a test command in .claude/CLAUDE.md. Use for thorough guideline alignment or testability cleanup before /optimus:unit-test.

Refactors existing code for guideline compliance and testability using 4 parallel analysis agents (guideline compliance, testability barriers, duplication/consistency, code-simplifier). Two goals — align code with project guidelines AND make untestable code testable so /optimus:unit-test can safely increase coverage. Use after /optimus:init to align existing code, before /optimus:unit-test to remove testability barriers, or periodically to prevent tech debt. Supports "testability" focus (after unit-test flags untestable code) or "guidelines" focus (after init establishes rules) to prioritize finding categories, and flexible scoping. For iterative refactor in a loop, use `/optimus:refactor-deep`.

Removes files installed by /optimus:init and /optimus:permissions from the project. Compares each file against plugin templates and classifies as unmodified, likely generated, or user-modified. Always asks before deleting. Git-tracked files are noted as recoverable. Tests are never touched. Monorepo and multi-repo aware. Use for clean reinstall or to stop using optimus.

Use when starting a new project or product and you want a docs-first plan before writing code — scaffolds an empty, product-neutral spec-driven-development cascade (product vision, MVP PRD, target tech-stack) for a human to fill, then hands off to brainstorm. Emits skeletons only; it never authors product content and never overwrites existing docs.

Guides test-driven development — decompose a feature or bug fix into behaviors, then cycle through Red (failing test) → Green (minimal implementation) → Refactor for each one. Requires /optimus:init and working test infrastructure. Use when starting a new feature or bug fix with test-first discipline.

Iterative test-coverage improvement loop — dispatches `/optimus:unit-test` (unit-test phase) and `/optimus:refactor` with testability focus (refactor phase) into fresh subagent contexts per cycle, applies tests, runs the test suite, bisects refactor failures, and continues until coverage plateaus or the cycle cap (default 5, hard cap 10). Use to drive coverage up on a codebase that has untestable barriers — the loop alternates between writing tests and unblocking testability so a single skill cannot stall.

Implements an approved spec by having Claude design and run its own Claude Code dynamic workflow (real parallel subagents) — you hand it the goal and constraints, it chooses the orchestration. Test-first is enforced as a quality bar (tests accompany or precede code and the suite is left green), not as supervised Red-Green-Refactor. A peer of /optimus:tdd for spec implementation; prefer it for large or parallelizable specs where one linear pass is slow. Requires /optimus:init and a spec (auto-detects docs/specs/ or docs/jira/, or pass a path). Uses meaningfully more tokens than a normal session. Use when a spec is ready to build and you want fan-out implementation instead of turn-by-turn TDD cycles.

Creates a git worktree for isolated parallel development — new branch in a separate directory with project setup and test baseline. Enables multiple Claude Code sessions on different tasks simultaneously. Multi-repo aware. Use when you need to work on something else without disturbing current work.

>

Trust infrastructure for agents that act. Verify identity and scoped authority, reuse trust receipts, and protect secrets across IDEs and runtimes.

>

>

Use when answering questions with STRING database MCP tools for protein-protein interactions, interaction partners, network images, interactive STRING links, evidence pages, functional enrichment, PPI enrichment, functional annotations, proteins associated with terms, homologs, sequence search, species lookup, or STRING limitations.

Pseudonymises medical and clinical documents by replacing patient identifiers with labelled tokens (e.g. [PATIENT_NAME_1], [NHS_NUMBER_1], [DATE_OF_BIRTH_1]) so the text can be safely processed by AI or shared, with clinical meaning intact. Combines a deterministic pattern layer (NHS numbers with Modulus-11 validation, UK National Insurance numbers, dates of birth, UK postcodes, phone numbers, emails, hospital/MRN numbers) with contextual reasoning for patient names, postal addresses and identifying ages, then returns the redacted document plus a redaction report. Use when the user wants to redact, de-identify, anonymise or pseudonymise a medical letter, clinical note, discharge summary, referral or patient record, or before pasting clinical text into another AI tool. Can also re-identify (reverse the redaction) by restoring original values from a token map, and offers a stricter HIPAA Safe Harbor mode for US de-identification (all dates, ages, and the remaining HIPAA identifiers).

>

Swap or bridge tokens through Ophis, an intent-based DEX aggregator. Parse a natural-language request, get a best-execution quote, build an EIP-712 order, sign it in the user's own wallet, and submit it. Non-custodial: the agent never holds keys or funds.

>

Professional frontend standards for building, scaffolding, extending, or reviewing any UI or frontend project — new or existing — even when standards aren't explicitly asked for. Keeps generated code consistent, reusable, secure, and production-quality. Framework-agnostic: React, Vue, Angular, Svelte, plain JS.

>

>

Look up concert setlists and live-music history via setlist.fm. Use when the user asks what songs an artist played at a show, their tour setlists, what was performed at a venue or on a date, or wants to find concerts by artist, venue, city, or year. Triggers on phrases like "what did Radiohead play at...", "Phish setlist for...", "shows at Red Rocks", "what songs were played on this tour", or any request about concert setlists, gigs, tours, or live performances. Requires setlist-mcp installed and the setlist server registered (see Setup below).

Use throughout long sessions to keep the context window small — archive large tool outputs to LETHE and recall them on demand, saving tokens.

Search Etix events, venues, and performers and pull event/venue details via MCP. Triggers on phrases like "find events on etix", "etix tickets for", "what's playing at <venue> on etix", "etix event details for", "search etix for <artist>", or any request involving Etix events, venues, performers, or showtimes. Requires etix-mcp installed and the fetchproxy extension active with an open etix.com tab (see Setup below).

>-

>

Generate and edit images with Google Gemini image models via MCP. Use when the user asks to generate, create, or edit images using Gemini or Nano Banana models, wants to produce a consistent set of images from a prompt, or needs to compose/blend multiple images. Triggers on phrases like "generate an image of", "edit this image with Gemini", "create a set of consistent images", "use Nano Banana to make", or any request to produce images via the Gemini image API. Requires the @chrischall/gemini-mcp package installed and the gemini server registered (see Setup below).

>-

Get a Budgetary pre-flight token-spend estimate for a task before running it. Use whenever the user wants to know how many tokens a task is likely to consume before doing it.

Bootstrap a brand-new Avo tracking plan from scratch through the Avo MCP. Use when the workspace is empty or near-empty and the team is deciding what to track — runs a purpose meeting (problems, goals, metrics, key funnels), proposes a naming convention (structure/casing/tense), a starting set of categories, and start-milestone-complete events with event/user/system properties and constraints, then creates a branch. Triggers on "we're new to Avo", "where do we start", "set up tracking from scratch", "what should we track", "bootstrap a tracking plan", "design our first tracking plan". If the workspace already has a substantial plan, use the data-designer skill instead.

Work with an existing Avo tracking plan through the Avo MCP. Use to explore and search events, properties, and metrics; look items up by exact name or id; list items by structural filters; review and prioritize branches; design tracking for a new feature or PRD against the existing plan; make small branch edits (rename, allowed values, attach properties); implement a branch's changes in source code; and combine Avo with a data MCP (Amplitude, Mixpanel, PostHog, BigQuery, Snowflake, Databricks, Redshift) to diagnose tracking gaps. Triggers on "what events do we have for", "how is X tracked", "add tracking for", "design tracking for", "review the X branch", "what's on branch X", "implement the X branch for [source]", "look up [event/property/metric]", and analysis prompts like "where are users dropping off" or "how is [funnel] performing". If the workspace is empty or brand-new, use the data-designer-new-plan skill instead.