All activity
AI writes code. The hard part is everything around it — task selection, branch management, review, CI, PR comments, cleanup, and shipping.
AgentSys orchestrates it all: 14 plugins, 43 agents, 30 skills. Give it one task and it autonomously routes exploration → planning → implementation → review → CI → merge. Zero babysitting.
Built on agnix-validated configs. Works across Claude Code, OpenCode, Codex CLI, Cursor, and Kiro. 3,750 tests. Production-grade.
AgentSysAI writes code. AgentSys automates everything else.
Skills don't trigger. Hooks fail silently. Configs conflict. You find out only when the agent breaks.
agnix catches all of it before it hurts.
Try now (no install): npx agnix .
npm install -g agnix | cargo install agnix-cli | brew install agnix
270 rules: Claude Code, Cursor, Codex, OpenCode, Kiro, Copilot, Cline, Gemini CLI
Validates CLAUDE.md, SKILL.md, hooks, MCP configs
67 auto-fixable (--fix) · LSP: VS Code, JetBrains, Neovim, Zed
GitHub Action + browser playground · Free
agnixThe missing linter for AI coding tools configs
Avi Feneshleft a comment
Hey PH 👋 I'm Avi — I ship Valkey/GLIDE + AWS by day and live inside AI coding tools (Claude Code, Cursor, Codex, OpenCode, Kiro) by night. After one too many wasted evenings debugging why a skill wasn't triggering, a hook was silently failing, or a CLAUDE.md was conflicting with my Cursor rules — I stopped guessing and built the tool I actually needed. agnix came out of that. 270 rules sourced...
agnixThe missing linter for AI coding tools configs
Avi Feneshleft a comment
Hey PH! I'm Avi — I ship production code on Valkey/GLIDE + AWS by day and live inside AI coding tools by night. I built AgentSys because AI can write code now, but everything *around* the code — picking the task, managing branches, multi-agent review, CI failures, PR comments, cleanup — still requires constant babysitting. AgentSys is the runtime that handles all of it. You type one task. It...
AgentSysAI writes code. AgentSys automates everything else.