alirezarezvani/claude-skills: what is in the 462-part plugin

What is alirezarezvani/claude-skills?

alirezarezvani/claude-skills is one of the largest single collections indexed on Flowy: 462 components in one repository, split into 114 agents, 146 commands, 200 skills, and 2 MCP servers. The GitHub repository itself carries 24,957 stars, a lot of validation for a bundle this size rather than one narrow tool. Flowy files it under the Agents category, though it reads more like a personal toolkit spanning several unrelated areas: a virtual C-suite of business advisors, a full engineering specialist bench, a Playwright testing kit, research and marketing tools, and everyday productivity workflows, all kept in one repository by one author.

Is there really a virtual C-suite in here?

Yes, and it is the single biggest idea in the collection. There is a named advisor agent for most executive seats, including cs-ceo-advisor, cs-cfo-advisor, cs-cto-advisor, cs-cmo-advisor, cs-coo-advisor, cs-ciso-advisor, cs-general-counsel-advisor, cs-caio-advisor for AI strategy decisions, cs-cdo-advisor for data strategy, and cs-vpe-advisor for engineering delivery and hiring. cs-chief-of-staff routes a question to the right one of these, boardroom runs a multi-role deliberation when a decision crosses several domains, and founder-mode is the single-command entry point for when you are not sure which advisor fits. Most seats also carry a matching interrogation skill, such as cfo-review or cto-review, that grills a plan before you commit money or headcount to it.

What does it give engineers day to day?

A specialist skill for most of a stack: senior-frontend, senior-backend, senior-fullstack, senior-devops, senior-secops, and senior-architect cover the general roles, while database-designer, api-design-reviewer, aws-solution-architect, azure-cloud-architect, gcp-cloud-architect, docker-development, kubernetes-operator, and helm-chart-builder cover narrower ground. The agents cs-fullstack-engineer, cs-backend-engineer, and cs-frontend-engineer each open with a set of forcing questions about load, tenancy, and data sensitivity before picking a stack profile and handing off to specialists like api-design-reviewer and database-designer. For a diff-level check, karpathy-coder and karpathy-reviewer apply the same four coding principles, the second reachable directly through the karpathy-check command.

Is there a dedicated testing toolkit?

Yes, a Playwright-focused set. pw is the general toolkit and pw-init sets up a new project. generate writes new tests from a component or route, fix repairs a flaky one, and migrate moves a suite over from Cypress or Selenium, with the agent migration-planner mapping the work file by file first. coverage reads Istanbul or LCOV output for gaps, and report renders the results. The agent test-architect plans strategy for a complex app, tester runs the suite, and test-debugger diagnoses a failure in depth. browserstack and testrail connect out to those services directly.

What about research, content, and marketing?

cs-deep-research, cs-dossier, cs-patent, cs-grants, cs-litreview, and cs-pulse each interview you before searching, then triangulate an answer across several independent sources. The agents content-strategist and cs-content-creator build and grade long-form content, cs-aeo audits a page for the signals that get it cited by an AI answer rather than ranked by a search engine, and seo-auditor runs a more traditional technical pass over a README or docs folder. A separate set handles LinkedIn specifically: cs-linkedin routes the request, and cs-linkedin-post, cs-linkedin-profile, cs-linkedin-outreach, cs-linkedin-analyze, and cs-linkedin-repurpose each cover one stage of writing, auditing, or reusing a post.

Anything for everyday productivity or a team's own knowledge base?

cs-deep-work and cs-time-block build a time-blocked schedule around a hard deep-work ceiling, cs-meeting-prep prices a meeting in real dollars before it gets scheduled, cs-inbox-triage classifies email against a knowledge base that cs-inbox-setup builds first, and cs-weekly-review runs a full GTD-style review. Separately, llm-wiki, wiki-ingestor, wiki-librarian, and wiki-linter maintain a personal or team knowledge vault: ingesting a source, answering questions with citations, and checking for broken links and stale pages.

What about compliance and audit prep?

A named advisor agent per framework: cs-soc2-auditor, cs-ciso-iso27001, cs-cqm-iso13485, cs-fda-qsr-auditor, cs-dpo-gdpr, cs-ai-act-compliance, and cs-aims-iso42001, each coordinating with the frameworks it overlaps instead of treating a SOC 2 review and an ISO 27001 review as unrelated work. compliance-os sits above all of them, mapping which frameworks apply to a given company profile and computing the overlap between them. soc2-audit-prep, iso27001-audit-prep, and gdpr-audit-prep run the same interrogation on demand, outside the full compliance-os flow.

What do the two MCP servers do, and how do I install any of this?

The pack lists two MCP servers, atlassian and tessl, without a description for either. Other components reference Jira and Confluence work directly, which points at what the Atlassian one is for, but tessl's purpose is not stated anywhere Flowy can read. Flowy indexes this plugin; it does not host, bundle, or maintain it. Install it from the listing page linked at the top of this guide, which shows the current install command, and every component stays on the author's own release cadence.