vaayneCommunity·14 Agent Skills Included

agent-kit

Plan-first coding workflow with review, refactoring, and handoff

Runs a disciplined development lifecycle: scout unknowns, stress-test designs, write decision-first plans, implement phase by phase, and review code with adversarial multi-agent checks. Eliminates vague prompts, skipped verification, and false-positive review noise. Also covers cloud file management, Google Workspace and Lark operations, email sending, and prose cleanup.

52|4

↗

npx skills add vaayne/agent-kit --all -g -y

All Skills in This Repository (14)

Pure Emerald Level Indicators

📦 In Repo

vaayne

kreuzberg

Extract text, tables, metadata, and images from 91+ document formats.

specs-dev

Coordinate plan-first development workflows with gate-based reviews and handoffs.

handoff

Generate a handoff.md document capturing goals, progress, decisions, files changed, blockers, and next steps.

gws

Control Google Workspace services from a single CLI.

lark

Orchestrate Feishu/Lark CLI modules to automate cross-module workflows.

mori-agent-bridge

Coordinate agent-to-agent messaging and workflows across Mori panes.

tap-web

Automate web lookup, page reading, and content extraction via the tap CLI.

reddit-posts

Generate subreddit-aware Reddit post titles, bodies, and comments.

python-script

Automate Python tasks with structured logging, validation, and safety checks.

vertex-ai-image

Generate, edit, and describe images using Google Gemini models via CLI.

code-simplifier

Refactor code while preserving functionality and enforcing project-wide standards.

twitter-posts

Generate publish-ready X post copy and threads from user briefs.

Community

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install agent-kit?▼

Run `npx skills add vaayne/agent-kit --all -g -y` in your terminal to install all skills in this collection globally.

What is the spec-dev workflow?▼

It chains grill (stress-test the idea), blueprint (write a decision-first plan.md), and mason (implement phase by phase with commits and handoffs), with optional scout and code-review steps.

How does the code-review skill avoid false positives?▼

It spawns four parallel reviewers (bug hunter, security auditor, architecture critic, correctness prover), then uses fresh verifier subagents to refute each finding against the actual code before reporting it.

Does agent-kit work with Claude Code and other runtimes?▼

Yes. All skills follow the standard SKILL.md format and sync into Claude Code, Pi, Codex, and other compatible runtimes.

Can it manage cloud storage and Google Workspace?▼

Yes. It includes skills for OpenList/AList file management, Google Workspace operations via the gws CLI, Lark/Feishu via lark-cli, and sending email through Cloudflare.