skills

Review code for overengineering and avoidable complexity, then write an HTML findings report. Use when the user asks whether code is too abstract, too layered, too generic, prematurely extensible, hard to follow, or more complex than the problem requires.

Use when implementing, changing, refactoring, reviewing, or accepting any code, config, build, migration, script, UI, API, dependency, or test change that needs strict executable proof instead of grep, source-text checks, snapshots-only assertions, import-only checks, or manual verification.

Audit and improve GitHub repository hygiene using evidence from local files, package manifests, dependency metadata, and GitHub settings. Use when the user asks for best practices, repo cleanup, repository hygiene, project health, package.json or dependency hygiene, monorepo health, npm/package publishing readiness, GitHub setup, maintainer readiness, contribution workflow, security posture, CI checks, issue or PR templates, branch protection, Dependabot, CodeQL, licensing, release readiness, or open-source/private repository governance.

Explain confusing ideas in extremely plain language, using simple references, analogies, examples, and step-by-step breakdowns. Use when the user asks to dumb something down, simplify, explain like they are new to the topic, make a problem easier to understand, turn dense writing into plain English, unpack jargon, or create beginner-friendly explanations for technical, business, academic, legal, medical, financial, or code-related material.

Research and implement a code change from an existing issue, ticket, or issue-like artifact. Use when the user says "$from-issue", asks Codex to work from a GitHub issue, local markdown issue, tracker ticket, bug report, feature issue, or AFK implementation ticket, and expects codebase research, linked PR verification, blast-radius analysis, targeted questions, a patch, verification, branch push, and pull request creation.

Slowly walk codebases and related project evidence to diagram data flow, workflows, dependencies, integrations, and connections between systems; record findings; verify each edge; and recommend fixes for breaks, edge cases, missing links, and disconnects. Use when the user asks to create, audit, validate, or troubleshoot architecture diagrams, workflow diagrams, integration maps, service maps, data lineage diagrams, sequence diagrams, system dependency maps, or cross-system debugging reports.

Review code changes for avoidable tech debt through function scope, naming, and file organization. Use when the user wants to know whether a change fits the existing codebase, introduces shallow abstractions, uses misleading names, or places code in the wrong module or directory.

Turn an architecture discussion, design conversation, rejected proposal, or selected refactoring direction into a complete Architecture Decision Record. Use after a conversation has clarified context, decision, alternatives, tradeoffs, consequences, or an architecture choice that future contributors should not re-litigate.

Audit a project, repo, app, or site by assuming its documented claims are false until disproven through code inspection, tests, builds, runtime checks, and contradiction hunting. Use when the user wants an adversarial validation that attempts to disprove README, ADR, PRD, roadmap, or marketing claims before accepting the project as working.

Verify whether a project, repository, or app does what it claims by inspecting the full codebase, running available checks, exercising behavior when possible, and producing an evidence-backed completion report with gaps and next steps. Use when the user asks to audit project completion, validate implementation against stated goals, or determine what remains to finish a repo or app.

Research the current project and recommend new or next features based on repository context, product surface, docs, issues, tests, TODOs, recent work, and user goals. Use when the user asks what to build next, wants feature recommendations, or needs a prioritized roadmap from the current codebase.

Create a single post-mortem GitHub issue and pull request from completed work in the current session. Use when the user wants to turn the contribution just completed into a tracking issue, commit or push the work, and open a PR without decomposing it into multiple implementation tickets.

Scaffold agent-ready project documentation for a new or existing repository, including AGENTS.md, CONTRIBUTING.md, DESIGN.md, issue templates, and pull request templates. Use when a user wants to initialize project docs for coding agents, contributors, design decisions, GitHub issues, or PR review workflows.