Claude Skills by LFTPadilla

> Step-by-step guide for adding a skill to this kit in all 4 agent formats. From a 5-minute skill to a 30-minute one.

Universal Harness Adapter — Delegate bounded coding-agent work across local agent harnesses (Pi, OpenCode, Codex CLI, Claude Code CLI, DHS). Use for explicit cross-harness requests, model comparisons, external-only runtimes, or named models requiring a CLI fallback. Do not use for generic requests to spawn subagents already exposed by the current session.

Operate as an unsupervised, long-running task with no user questions after the user explicitly invokes that mode. Run a mandatory Plan + Research phase before execution, keep a journal and checkpoints, produce a final report, and only run shutdown or production steps when explicitly authorized in the same session.

Audit, scaffold, or refactor any software repository into the Agent-Native Repository Architecture (ANRS-1.0) with Hub-and-Spoke context, O(1) REGISTRY.yaml, and shallow directory ergonomics.

Renders network diagrams, flowcharts, and infrastructure diagrams as PNG images using SVG generation and sharp. Use when the agent needs to output a network diagram, system architecture, flowchart, or infrastructure map as a viewable image. Triggers on: "make a diagram", "render this network", "show infrastructure as image", "draw topology", "create flowchart", "generate diagram".

Use when the user requests diagrams, flowcharts, architecture diagrams, ER diagrams, UML / sequence / class diagrams, network topology, ML/DL model figures (Transformer/CNN/LSTM), mind maps, or any visualization. Also use proactively when explaining systems with 3+ components, complex data flows, or relationships that benefit from visual representation. Best suited when the diagram needs custom styling, rich shape vocabulary, swimlanes, or exportable images (PNG/SVG/PDF/JPG). Generates .drawi...

Create Microsoft Excel .xlsx spreadsheets from tables, CSV/TSV, Markdown tables, or JSON data. Use when the user asks for Excel, spreadsheet, .xlsx, tabla editable, budget, inventory, report table, calculations, or data they need to open/edit in Microsoft Excel.

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

Essential Git commands and workflows for version control, branching, and collaboration.

Control Herdr, a terminal multiplexer for coding agents. Use only when the user explicitly mentions Herdr or asks to use Herdr to inspect or control panes, tabs, workspaces, commands, or another agent. Do not use merely because a task could benefit from a background terminal, delegation, or parallel work. Requires HERDR_ENV=1.

Write in a direct, human, practical style with short sentences, active voice, and strict phrasing bans.

Two-stage image generation workflow. First asks clarifying questions, then uses minimax for fast iterative drafts, then gemini-3-pro-image-preview for final polish. Trigger on image generation requests, editing/refining images, or when user wants to produce a final version.

Survey any codebase as a senior advisor and produce prioritized, self-contained implementation plans for OTHER models/agents to execute. Strictly read-only on source code — never implements, fixes, or refactors anything itself. Use when asked to audit a codebase, find improvement opportunities (bugs, security, performance, test coverage, tech debt, migrations, DX), suggest features or where to take the project next (roadmap, product direction), or generate handoff plans for another agent to i...

Develop, debug, test, or review Java/JVM repositories without guessing their stack. Use for Java, Maven, Gradle, JUnit, Mockito, Testcontainers, Spring Boot, JVM toolchains, build failures, test failures, coverage, or Java CI work. Discovers wrappers and pinned JDK settings first, preserves project conventions, uses focused-to-broad verification, and reports real build evidence.

Find dead code, unused files, unused exports, and unused dependencies in a JavaScript/TypeScript project using knip. Use before a refactor, before a PR, or when the user asks to clean up bloat, remove unused deps, or shrink the codebase.

Exploratory QA of a running app by driving a real browser like a user via the Playwright MCP server. Use when the user wants to "QA" a feature live, walk a flow as a real user, smoke-test before a PR, or find issues that scripted specs miss. Complements deterministic Playwright specs — this is the human-like exploration layer.

Explicit orchestrator mode for Antigravity, Codex, Claude Code, PI, OpenCode, or Hermes. Plans, decomposes, chooses worker models via LiteLLM/SSoT, delegates execution (headless by default, Herdr/tmux on visual request), enforces terminal layout limits, and independently verifies results on disk.

Work with PDF files using the environment's built-in PDF and document tools. Use when the user wants to inspect, summarize, split, merge, compress, or convert PDFs.

Make a Playwright E2E suite stable and realistic — kill flaky tests and authenticate like a real user via storageState (login once, reuse). Use when E2E tests are flaky, slow, re-login in every test, mock auth instead of using it, or when hardening a suite before relying on it.

A pattern → severity → fix checklist for reviewing security-sensitive code (auth, payments, user input, file uploads, webhooks). Use as the LLM-reasoning complement to a deterministic SAST scan — semgrep catches known patterns, this catches the contextual ones. Run on any diff touching a trust boundary.

Run static analysis (SAST) over the codebase with semgrep to find real bug and security patterns — injection, auth gaps, secret handling, dangerous APIs. Use before a security-sensitive PR, on auth/payment/data-handling code, or when the user asks for a deterministic security scan.

Write self-healing browser automation with Stagehand (act/extract/observe) — natural-language steps on top of Playwright that survive UI changes without rewriting selectors. Use for the parts of a flow that change often, or when scripted Playwright selectors keep breaking. Pilot it against one volatile flow before adopting widely.

Acts as an engineering Tech Lead — mentoring the developer across 'learning', 'flow', and 'autonomous' modes, managing cognitive debt, and enforcing architectural review gates while delegating execution to the `orchestrate` skill. Use when the user says "$tech-lead", "tech lead", asks for architectural mentorship, or needs tech lead guidance on a project.

Renders LaTeX math to PNG, JPEG, WebP, or AVIF images using MathJax (TeX→SVG) and @svg-fns/svg2img. Invoke whenever the agent needs to output LaTeX as a viewable image (equations, formulas, notation).

Delegate tasks to coding agents running in tmux panes across multiple harnesses (Codex, Claude Code, Pi, OpenCode, cursor-agent). Covers pane allocation, reliable multiline buffer injection, harness-specific completion detection, queued inputs, and disk-level diff auditing.

Navigate websites and interact with web pages as a human user. Use when needing to browse, scrape, extract data from, or interact with websites that require real browser behavior (login flows, dynamic content, CAPTCHAs, bot detection, etc.). Triggers on phrases like "browse to", "go to website", "open webpage", "check the page", "extract from", "login to", "navigate to", or any task requiring real web interaction with a browser.

Create Microsoft Word .docx documents from a title and body text. Use when the user asks for a Word file, .docx, formal letter, proposal, quote, report, or document to edit/share in Microsoft Word.