Personal AI assistant: screenpipe→Obsidian ingest + requirement radar → approval cards → autonomous claude execution. macOS menu-bar app + Python pipeline.
This repository also configures its own agents. See what zelin-ai-assistant tells them →
Skill Claude CodeCodex
A workflow for processing files from an `unprocessed/` folder, including screen recordings, exported conversations, notes, PDFs, and images. It moves information through raw and wiki stages.
Skill Claude CodeCodex
Read Zelin's AI-assistant board and file captures or progress comments with boardctl. Use for card IDs (R-xx/MS-xx), progress notes on executing work, or proposing follow-up candidates — never for approving, accepting, or moving cards.
Skill Claude CodeCodex
ASK mechanics: Claude Code → AskUserQuestion with one single-select (tier) and one multi-select (checks). Other harnesses → numbered text menu; wait. Headless (agent finishing work) → --chosen-by headless; the report records recommended, not confirmed. You may add checks, never silently drop core ones: an unticked…
Skill Claude CodeCodex
ASK mechanics: Claude Code → AskUserQuestion (one single-select tier, one multi-select checks). Other harnesses → numbered menu; wait. Headless (agent finishing UI work) → --chosen-by headless, standing default tier 2 --screens changed; the report records recommended, not confirmed. The reference is a CLI input, never…
Skill Claude CodeCodex
Write and edit the owner's formal English output (emails, papers, articles, docs, reports, CV/resume) in a precision-first style modeled on Munindar P. Singh's editorial standards. Use whenever drafting or revising any email, academic paper, blog article, README, report, professional message, or CV/resume. Also use…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: