The source for the systems on this site, read straight from the repositories.
01 / 08mattrguerra/meridian-ops-board
An operations dashboard and daily AI briefing for a print and display shop: orders, CRM, to-dos, meetings and shared inboxes in one page per person. The company and its data are fictional.
02 / 08mattrguerra/meridian-crm-mcp
A local MCP server that lets Claude read and write a Bitrix24 CRM, with tasks, telephony and website chats: 30 tools, writes locked to named funnels, redacted logs. The company is fictional.
03 / 08mattrguerra/meridian-orders-mcp
A read-only MCP server that lets Claude look up orders, customers and statuses, with a four-layer write guard and a shared rate limiter on Cloudflare Workers. The company and its order system are fictional.
04 / 08mattrguerra/meridian-skills
Twenty Claude skills for the weekly work of a print and display shop: CRM routing, outreach enrichment, storefront data, artwork prep, SOPs and customer email in one voice. The company is fictional.
05 / 08mattrguerra/meridian-outreach-vault
The Obsidian vault behind a print and display shop's quarterly client outreach: a weekly pipeline that keeps a partner-client census current, and audited email drafts a person approves. The company and its clients are fictional; the client book is a generated sample run through the real pipeline.
06 / 08mattrguerra/better-memory-pipeline
A nightly pipeline that turns a personal Obsidian knowledge base into memory any Claude can recall from: a fail-closed gate walls off private material and stages only what changed, and a Claude skill distills it into a Notion knowledge graph. Runs on a sample vault.
07 / 08mattrguerra/transcription-pipeline
A watch-folder pipeline that turns recordings into transcripts that say who said what: local Whisper by default or a hosted engine for quality, word-level alignment, pyannote speaker separation, and a copy filed into a vault inbox. The manual transcriber it grew out of ships alongside.
08 / 08mattrguerra/cortex
A native desktop client for working with models directly, in Tauri and React, built so the price of every reply is in front of you: live per-model prices, the cost of each answer, a monthly budget, and an estimate before you send. Streaming with stop, a global quick prompt, search, presets, and export to an Obsidian vault.