Skills, agents & plugins

The tools I build to do agentic software engineering — Claude Code skills that enforce standards, agents that run real workflows, and plugins that wire local models and external systems into the loop. Filter by type, or browse the whole set.


6

Total

3

Skills

2

Agents

1

Plugins

The full set

Each entry is something I run on real work. Published items link to a repo or details; coming-soon items are in active development.

pooya-llm-aeo-revise

Rewrite copy so it ranks for SEO and gets cited by AI answer engines

Skill

A Claude Code skill that revises text for or about Pooya Golchian and pooyagolchian.com so it reads as human, ranks in traditional search, and earns citations from ChatGPT, Perplexity, Google AI Overviews, and Claude. It strips AI-isms, enforces the brand voice, and tightens LLM-facing content.

  • Rewrites blog intros, About copy, service pages, and case studies for AEO
  • Cleans up meta descriptions, OG snippets, and FAQ blocks
  • Enforces the Pooya voice and strips filler AI phrasing
  • Tunes llms.txt entries and other LLM-facing content for citation
AEOSEOClaude CodeCopywritingLLM

seo-audit

On-demand technical and on-page SEO diagnosis for a URL

Skill

A Claude Code skill that audits a page across technical SEO, on-page elements, content quality, and E-E-A-T against a structured checklist, then reports concrete fixes. Built for diagnosing why a page underperforms in search.

  • Runs a multi-point technical and on-page SEO checklist on a URL
  • Flags content-quality and E-E-A-T gaps with specific fixes
  • Detects site type and selects the right pages to inspect
  • Produces a prioritized, actionable report
SEOAuditClaude CodeE-E-A-TTechnical SEO

next-best-practices

Next.js conventions enforced while you write and review code

Skill

A Claude Code skill that applies current Next.js best practices — file conventions, RSC boundaries, data patterns, async APIs, metadata, error handling, route handlers, and image/font optimization — when writing, reviewing, or refactoring App Router code.

  • Enforces App Router file conventions and server/client boundaries
  • Applies correct data-fetching and async API patterns
  • Guides metadata, error handling, and route-handler design
  • Catches image, font, and bundling anti-patterns in review
Next.jsReactClaude CodeApp RouterBest Practices

use-local-llm

React hooks for streaming local LLM responses

Plugin

React hooks for streaming responses from local LLMs — Ollama, LM Studio, llama.cpp, and any OpenAI-compatible endpoint. Direct browser-to-localhost streaming with no server, no API routes, and zero runtime dependencies beyond React.

  • Streams from Ollama, LM Studio, llama.cpp, and OpenAI-compatible endpoints
  • Runs direct browser to localhost with no server or API routes
  • Manages message history, abort, clear, and error handling
  • Ships at 2.8 KB gzipped with zero runtime deps beyond React
ReactTypeScriptLLMOllamaStreaming
$ pnpm add use-local-llm

RevOps AI

AI sales agent built on Notion through the Model Context Protocol

Agent

A Revenue Operations platform where Notion is the database and Gemini 2.5 Flash is the AI sales agent. Kanban pipeline, AI lead scoring, pre-call briefings, and natural-language deal management — all driven by 22 Notion operations exposed through MCP.

  • Chains 22 Notion MCP operations per request autonomously
  • Scores leads 0–100 with written reasoning, saved back to Notion
  • Syncs a drag-and-drop Kanban pipeline to Notion in real time
  • Answers pipeline queries and drafts emails in natural language
AINotionMCPGeminiRevOps

AIDLC Orchestrator

An agent that runs a project through the eight AIDLC phases

AgentComing soon

A Claude Code agent that drives a build through the AI Development Life Cycle — framing the problem, writing the spec, scaffolding the eval harness, generating code, and gating every change on evals. In active development.

  • Walks a project through Frame, Spec, Scaffold, and Generate
  • Wires an eval harness in before feature code exists
  • Gates each generated diff on the eval suite
  • Keeps the build inside the boundary defined in the spec
AIDLCClaude CodeAgentEvalsOrchestration

Want a skill or agent built for your stack?

I build Claude Code skills, agents, and MCP integrations as part of the AIDLC method. Tell me what you want automated.

Get practical AI and engineering playbooks

Weekly field notes on private AI, automation, and high-performance Next.js builds. Each edition is concise, implementation-ready, and tested in production work.

Open full subscription page

Get the latest insights on AI and full-stack development.