raiderrobert

Use when reviewing a codebase's structural organization, not a PR or branch. Triggers on: codebase audit, structural review, architecture audit, code organization review, audit this codebase, how is this repo organized, code health check, audit the code. Produces a findings report against a standard checklist plus project-specific conventions.

Use when working on multiple tickets simultaneously, dispatching parallel subagents across worktrees, or needing a CI-fix loop to get branches green. Covers worktree setup, parallel planning, implementation, PRs, tests, and CI monitoring.

Use when a task has a clear implementation plan and should be executed by a subagent in an isolated git worktree. Triggers on phrases like: dispatch this, send to a subagent, work on this in a worktree, dispatch to worktree, background task, isolated execution.

Use when the user wants to review existing work on a branch, find issues, and dispatch parallel fixes that get cherry-picked back. Triggers include: review and fix, review and fix this branch, find and fix issues, dispatch fixes, parallel fix, review then fix, fix all issues, code review with fixes.

Use when code needs repeated review-fix cycles until clean. Dispatches a reviewer agent, then a fixer agent, commits, and repeats until findings converge to zero or a cap is reached. Triggers on: iterate until clean, review fix loop, converge, repeated review, keep fixing until done, iterative review.

Use for ALL Rust questions including errors, design, patterns, and coding. Triggers on: Rust, cargo, rustc, Cargo.toml, compiler errors (E0382, E0597, E0277, E0308, E0499, E0502, E0506, E0507), ownership, borrowing, lifetimes, traits, generics, async/await, tokio, concurrency, Result, Option, Error, thiserror, anyhow, Box, Rc, Arc, RefCell, performance, profiling, clippy, code review, anti-pattern, idiomatic Rust, code style, best practice

Use after generating any visual design, frontend code, or .pen file to check for generic AI defaults. Triggers on: review design, audit design, check design quality, design review, does this look good, is this generic, AI slop, design check, visual QA, before shipping design, review my frontend, check my CSS, quality check, polish, improve design. Catches common AI aesthetic convergence patterns.

Use when building web components, pages, or applications with HTML, CSS, JavaScript, React, Vue, Svelte, or any frontend framework. Triggers on: build a page, create a component, design a UI, landing page, dashboard, settings panel, login page, signup form, pricing page, hero section, card component, navigation, sidebar, modal, frontend, web design, HTML, CSS, styled, Tailwind, web app, website. Ensures distinctive, production-grade output that avoids generic AI aesthetics.

Use when writing or reviewing a specification document — a normative definition of a format, protocol, interface, schema, or contract. Triggers: write a spec, SPEC.md, docs/spec/, RFC-style doc, formalize this design, define what X is, conformance, normative, this reads like marketing, make it more RFC-like, spec register. Ensures spec-grade register and conformance structure; prevents marketing prose, coined jargon, and speccing the status quo instead of the target.

Use when adding research papers, managing bibliography entries, or adding quotes to citation files. Ensures papers are cataloged consistently across bibliography, quotes, and review files so research is findable and citable. Triggers on: add paper, add citation, new source, add quote, update bibliography, cite this, catalog this paper, add to references, scribe.

Use when writing, creating, editing, improving, reviewing, or auditing Claude Code skills. Triggers on: write a skill, create a skill, new skill, skill template, skill structure, skill authoring, edit skill, improve skill, review skill, audit skill, skill quality, SKILL.md, skill frontmatter, skill description, skill triggers, make a skill for, how to write a good skill, audit until clean, keep auditing, iterate on skills, iterative skill audit, converge skill quality

Generate high-quality analysis, plans, or designs by running a structured bakeoff. N agents independently produce solutions in isolated worktrees, M critics review each from orthogonal angles, then synthesize the best elements into a composite. Use when the problem is ambiguous, high-stakes, or benefits from multiple perspectives. Triggers: bakeoff, diverge-converge, independent analysis, multi-perspective review, competitive analysis, adversarial review, stress-test a plan.

Use when the user has a list of claims, ideas, laws, quotes, or attributions and needs original sources and provenance for each. Triggers include: find references, find sources, source this, cite these, where does this come from, who said this, original source, provenance, attribution check, fact-check attributions.

Use when asked to review a branch, PR, or implementation where the quality of the approach itself is in question — not just code style, but whether the technique or design is sound. Also use when the user expresses suspicion about an approach and wants independent verification, or says: look up, research, what do others do, good practice, validate approach.

Use when you want to understand what reviewers actually enforce on a repo, discover undocumented team standards, onboard to a new codebase's culture, create a coding conventions skill from real review patterns, audit a mined skill, or verify that conventions are accurate. Trigger on "what do reviewers care about", "extract conventions from PRs", "mine review feedback", "what patterns does this team enforce", "create a conventions guide from PR history", "team standards from code reviews", "PR feedback analysis", "what does the team prefer", "audit this skill", "verify the conventions", "check if these quotes are real", "is this skill accurate".

Use when reviewing a project's public surface for implementation details that leaked through. Triggers on: leaky abstraction, implementation detail exposed, coupling to dependency, file extension choice, config key naming, API surface review, public interface audit, dependency name in user-facing surface, internal detail leaked, would this break if we swapped libraries

Search for, install, upgrade, and manage agent skills using the napoln package manager. Use when the user wants to find new capabilities, install a skill from a git repository, check for skill updates, view local modifications to installed skills, or resolve merge conflicts after upgrade.

First test skill in a multi-skill repo.

Second test skill in a multi-skill repo.

A valid test skill for unit testing.