18 repos · 14 plugins · 43 agents · 32 skills

Code does code work.
AI does AI work.

The ecosystem for AI agent tooling. Workflow orchestration, config linting, browser automation — each tool standalone, all composable.

18 Repos
14 Plugins
43 Agents
32 Skills
230 Lint Rules

Get Started

Pick a tool. Install in one command.

Workflow orchestration - 14 plugins, 43 agents, 32 skills.

$ npm install -g agentsys && agentsys

Or in Claude Code: /plugin marketplace add agent-sh/agentsys

Agent config linter - 230 rules across 10+ AI tools.

$ npx agnix .

Also: cargo install agnix or brew install agent-sh/agnix/agnix

Browser automation for AI agents - Playwright-based, no MCP overhead.

$ agentsys install web-ctl

Skills: web-browse (headless navigation) and web-auth (login handoff)

Built Different

Not another AI wrapper. Engineering-grade workflow automation.

Code does code work. AI does AI work.

Static analysis, regex, and AST for detection. LLMs only for synthesis and judgment. 77% fewer tokens than multi-agent approaches.

One agent, one job, done well

43 specialized agents, each with a narrow scope and clear success criteria. No agent tries to do everything.

Pipeline with gates

Each step must pass before the next begins. Can't push before review. Can't merge before CI. Hooks enforce it.

Validate plan and results, not every step

Approve the plan. See the results. The middle is automated. One approval unlocks autonomous execution.

Right model for the task

Opus for complex reasoning. Sonnet for validation. Haiku for mechanical execution. Quality compounds - early phases deserve the best model.

Composable, not monolithic

Every command works standalone. /deslop cleans code without /next-task. /ship merges PRs without the full workflow. Pieces compose but each piece is useful on its own.