One folder. Any agent. Any tool.
ACS defines a unified .agents/ folder for agent-ready projects. Write your context, skills, and permissions once — use them across Claude Code, Cursor, Copilot, and any future tool.
Agent configuration is fragmented. Your context, skills, and rules live in vendor-specific formats that don't travel.
ACS is additive. Start with context only. Add layers as your project grows.
What agents need to know. Architecture, conventions, do-not-touch rules.
Reusable capabilities. SKILL.md compatible with agentskills.io.
Named, single-shot tasks. Like slash commands for any agent.
Named subagents: reviewer, tester, docs-writer, and more.
Explicit allow/deny rules for file access. Committed to git.
Multi-step orchestration and lifecycle triggers.
ACS uses a three-tier loading strategy. Agents get the right context at the right time — without bloating the session.
Manifest, permissions, and project context load at session start. Skill names only — ~50 tokens each.
~200 tokens total overheadFull SKILL.md body loads when the agent decides the task matches. Commands and agents load on invocation.
< 5,000 tokens per skillReference files, scripts, and assets load only when the skill instructions reference them explicitly.
zero cost if unusedACS works alongside every existing standard. Add .agents/ to any project without breaking what's there.
Keep your AGENTS.md for broad compatibility. ACS adds structured layers on top. They live side by side in the same repo.
ACS skills use the exact same SKILL.md format. Any agentskills.io skill works inside .agents/skills/ with zero changes.
MCP handles runtime tools. ACS handles project configuration. Different layers, fully complementary.
Migrate incrementally. The planned acs compile CLI will generate CLAUDE.md from ACS automatically.
Create a .agents/ folder. Add acs.yaml. Write context. Done — any ACS-compatible agent picks it up automatically.