Claude Code Skills — the Licence-Safe Directory

Stages relevant changes, writes a conventional-commit message, and pushes — in one command.

Opens a pull request with a structured description drawn from the full commit range, not just the latest commit.

Runs a multi-angle review on an open PR: correctness, style, tests, security, and docs.

Sweeps a repo for OWASP-top-10 issues, hardcoded secrets, insecure dependencies, and auth logic gaps.

Generates unit tests for a specific function or module, with sensible edge cases.

Drafts a blog post, newsletter, or landing page from a brief — using the site's established voice.

Gathers sources, stats, frameworks, and competitive analysis for a topic — keeps research separate from writing.

Analyses existing content to extract the brand's voice patterns, then scores new drafts against them.

Writes high-conversion short copy — headlines, CTAs, subject lines, hero copy — using classic copywriting frameworks (PAS, AIDA, 4Ps).

Generates 10 headline variants against different frameworks (curiosity, benefit, controversy, listicle, how-to) and ranks them.

Drafts in a specific author's voice after reading 3-5 of their pieces — attribution-safe for transparency.

Turns one blog post into a newsletter + 5 social posts (LinkedIn, Twitter/X thread, Instagram caption) in one pass.

Rewrites a tweet or thread for the current Twitter/X algorithm preferences — engagement bait avoidance, read-time, link handling.

Runs a full on-page SEO audit on a URL or page file — title, meta, H1, schema, internal links, content gaps.

Scores a page for LLM-citability — direct answers, quote-ready structure, source attribution, FAQ coverage.

Generates JSON-LD schema markup for a page — Article, FAQPage, HowTo, Product, SoftwareApplication, BreadcrumbList.

Suggests contextual internal links for a page by scanning the rest of the site for topically-relevant anchors.

Rewrites meta title and description for a page, optimised for CTR on target query.

Builds the site, validates output, and pushes to Netlify — with rollback if the build fails post-deploy.

Checks the latest deploy status for all (or one) Netlify site — green/building/failed at a glance.

Spins up the project's dev server — handles npm install if needed, picks an available port, opens the browser.

Designs Helm charts, K8s manifests, and GitOps workflows — with sensible defaults for indie/small-team deploys.

Scrapes 3-5 competitor pages on the same topic, compares stance, and identifies content gaps + differentiation angles.

Summarises 3-5 academic papers on a topic into a plain-English brief with the core claim, method, and limitations of each.

Converts Word, PDF, or HTML documents into clean markdown — preserving structure, tables, and code blocks.

Analyses a meeting transcript for key decisions, action items, sentiment shifts, and unanswered questions.

Sweeps CTAs, URLs, or brand terms across a repo with tier-based policy — keep some, swap others, leave the rest.

Greps for residual outdated copy across a repo after a pivot — catches prose that escaped the CTA sweep.

Verifies a project's reference data (prices, models, deps) against upstream sources — flags staleness.

Automates the lifecycle: current → legacy → deprecated → retired as dates pass. Keeps model data in sync with vendor announcements.

Runs a battery of drift/regression checks on a site — CTA hygiene, data freshness, link health, cruft in pages/.

Verifies that a defined hub of pages all link to each other — catches missing cross-links that break SEO and UX.

Scaffolds a new Claude Code skill with proper frontmatter, invocation block, and SKILL.md spec compliance.

The canonical SKILL.md template from Anthropic's official skills repo — frontmatter fields, trigger syntax, invocation patterns.

Dry-runs a skill against a test input to verify trigger detection, output shape, and side-effect safety before shipping.

Audits a site for a legal disclaimer page + footer link; creates and wires up if missing.

Checks every dependency's licence — flags GPL contamination, commercial-incompatible licences, missing attribution.

Adds a new vendor to a vendor-directory site — creates the JSON data file, the detail page, the category listing, and cross-links.

Verifies a vendor's data against their actual website — pricing, features, contact details.

Presents a menu of your sites, loads per-site memory + recent git log, surfaces a session brief.

Audits GitHub Actions across all your repos for failures in the last 7 days — flags broken cron jobs, silent CI failures.

Launches 3-5 parallel QA subagents against a site, each with a distinct remit (SEO, monetisation, freshness, accuracy), synthesises findings.

Fires a smoke test at AI-powered API endpoints (Claude, GPT, custom agents) on live sites — flags 5xx, timeouts, broken auth.

Crawls a live site for 4xx/5xx responses — reports broken internal links and rotted outbound links.

Runs Lighthouse against homepages (or specific pages) across a portfolio — reports LCP, CLS, INP, and regressions.

Verifies the licence of a GitHub repo (or specific file) before you fork, republish, or include content from it.

Generates a human-readable changelog between two git tags — grouped by feat/fix/refactor, with author attribution.