Claude Skills by imisic

Commit the working tree with pre-flight checks, a drafted conventional message, and a sensitive-file gate.

Debug one bug by tracing it through your architecture's layers, or batch-apply a review report by severity.

Review, fix, and commit in one run, with a gate before each phase.

Shared review engine that a project's own review skill reads. Not run on its own.

Read this codebase and write, or sharpen, a review skill fitted to it.

Audit .claude/rules/ and CLAUDE.md against what the codebase actually does.

Turn recurring review findings into proposed rule and skill updates.

Pre-publish privacy gate. Runs before anything crosses from a private working environment to a public one: a plugin marketplace repo, a downloadable bundle, a public GitHub repo or gist, a shared skill or command. Scans the exact files that will ship for key material, credentials, internal hostnames, LAN IPs, personal home paths, work-email domains, secrets plumbing and internal or client project names with a deterministic scanner, then does a human pass for the contextual leaks a regex canno...

Optimize a web project for AI search visibility (GEO/AEO) so ChatGPT, Claude, Gemini, Perplexity and Copilot can cite it. Triggers: AI SEO, GEO, AEO, llms.txt, schema markup for AI, robots.txt for AI bots, structured data audit, WebMCP, agentic browsers, generative or answer engine optimization, "make my site AI-friendly", "optimize for ChatGPT", or any page whose AI discoverability matters even if unnamed. Also use when building a new site or page, to bake GEO in from the start. Owns machine...

Turn Search Console data, fetched over the API or exported by hand, into a ranked, evidence-backed action plan.

Finds what makes source code read as AI-written and points you at the parts that actually ship bugs. Sorts every tell into three buckets and fixes them in that order: bugs (swallowed errors, a made-up API, a left-in "rest of the code" stub), substance (tutorial-shaped boilerplate, over-engineering, code that ignores the surrounding repo), and cosmetic (emoji in source, a comment narrating the obvious, a generic process_data name, chat text pasted into a file). Ships a scanner for the surface ...

A canonical-voice and anti-slop layer for prose. Strips the cues that make writing read as AI-generated (the em dash, banned corporate/AI diction like delve, leverage, utilize, robust, paradigm, synergy, the "it's not just X, it's Y" cadence, sycophantic openers, bookending, leftover assistant boilerplate) and holds a deliberate, register- appropriate human voice instead. Ships a merged banned-vocabulary catalog and a scanner that gates on the hard bans. Use whenever writing, drafting, editin...

Finds the cues that make a website read as AI-generated and forces a deliberate, project- specific design choice instead of the model's default. It does not impose a look. It flags both the old default (shadcn/Tailwind stock theme, AI-purple, gradient hero text, gradient body text, unprompted neon glow, emoji-as-icons, the centered-hero-plus-three-cards skeleton) and the newer cream-plus-serif-plus-green "tasteful default" that anti-slop advice itself created, so you do not trade one default ...