Implements or fixes a specific, well-scoped GitHub issue in an isolated git worktree, validates it against the project's formatting/static-analysis gates, and leaves a log for human review. Trigger whenever the person asks to fix, implement, or pick up a specific issue by number (e.g. "fix issue #42", "implement issue 17", "can you handle issue #8"), or when an issue is tagged/labeled "Agent". Do not commit or open a PR — that stays with the human.
Performs a strict, two-phase static analysis of a codebase (frontend, then backend) to surface critical security vulnerabilities, bugs, dead code, and architectural anti-patterns. Use this whenever the user asks for a code review, audit, security review, vulnerability scan, bug hunt, or static analysis of any codebase or project, even if they don't use those exact words (e.g. "look over my code", "is this secure", "what's wrong with this project", "check my repo for issues"). Backend analysis is always graded strictly against official Laravel 13 documentation best practices; frontend analysis is framework-agnostic. Findings only, never code fixes, diffs, JSON schemas, or diagrams, only named issues with explanations of impact and severity.
Use this skill whenever Thomas has a folder of overlapping markdown notes (e.g. a /docs directory on a client project) that need to be condensed into one trustworthy summary before a client-side or team discussion. Trigger on phrases like "densify these notes", "check my docs for conflicting info", "clean up these notes before the client call", "some of these notes are wrong", "find where the team disagrees", or any request to consolidate scattered project documentation. The skill reads only markdown/text files — it deliberately skips MLD diagrams, PNGs, and other binary/image files since those aren't meant to be parsed as text. Whenever the same fact is stated with different values across files, the skill ALWAYS stops and writes a clarification/*.md file pinpointing exactly where the conflict lives, before producing anything else — it never silently guesses which value is correct. It then always produces a clean, densified summary doc, flagging any unresolved conflicts inline instead of papering over them. T
Security and code quality audit for traditional Laravel projects using Blade templates (no Inertia, no API layer). Triggers when Thomas asks to "audit", "review", "analyse", or "check" a Laravel+Blade codebase, or when he says things like "is this secure?", "any issues with my code?", "code review my Laravel project", "check for vulnerabilities", "best practices check", or pastes code and asks what's wrong. Also trigger when he mentions Sail, PSql, Blade files, or asks if his Laravel project is production-ready. Be pushy — if a Laravel + Blade project is being discussed and there's no recent audit, suggest running one.
Trigger when writing, reviewing, or debugging Laravel code. Enforces framework conventions: thin controllers, Form Requests for validation, Eloquent over raw queries, and proper use of built-in features. Also trigger on questions about Laravel architecture or "how do I do X in Laravel".
Use this skill whenever Thomas has a folder of overlapping markdown notes (e.g. a /docs directory on a client project) that need to be condensed into one trustworthy summary before a client-side or team discussion. Trigger on phrases like "densify these notes", "check my docs for conflicting info", "clean up these notes before the client call", "some of these notes are wrong", "find where the team disagrees", or any request to consolidate scattered project documentation. The skill reads only markdown/text files — it deliberately skips MLD diagrams, PNGs, and other binary/image files since those aren't meant to be parsed as text. Whenever the same fact is stated with different values across files, the skill ALWAYS stops and writes a clarification/*.md file pinpointing exactly where the conflict lives, before producing anything else — it never silently guesses which value is correct. It then always produces a clean, densified summary doc, flagging any unresolved conflicts inline instead of papering over them. T
Simplifies and refines PHP/Laravel code for clarity, consistency, and maintainability while preserving all functionality. Focuses on recently modified code unless instructed otherwise.