Agentic
Collaboration
Standard
One folder. Any ACS-compatible agent.
ACS defines a unified .agents/ folder for agent-ready projects. Instead of scattering project guidance across CLAUDE.md, .cursorrules, .github/copilot-instructions.md, GEMINI.md, .windsurfrules, .clinerules, .junie/guidelines.md, and a dozen more vendor-specific files, write your context, skills, and permissions once for ACS-compatible tools.
Every tool reinvents
the same wheel.
Agent configuration is fragmented. Your context, skills, and rules live in vendor-specific formats that don't travel. File examples below are illustrative and can evolve by vendor version.
Five layers.
One folder.
ACS is additive. Start with context only. Add layers as your project grows.
context/
What agents need to know. Architecture, conventions, do-not-touch rules.
skills/
Reusable capabilities. SKILL.md compatible with agentskills.io.
commands/
Named, single-shot tasks. Like slash commands for any agent.
agents/
Named subagents: reviewer, tester, docs-writer, and more.
permissions/
Explicit allow/deny rules for file access. Committed to git.
workflows, hooks, profiles
Multi-step orchestration and lifecycle triggers.
Progressive disclosure.
Load only what you need.
ACS uses a three-tier loading strategy. Agents get the right context at the right time — without bloating the session.
Always in context
Manifest, permissions, and project context load at session start. Skill names only — ~50 tokens each.
~200 tokens total overheadOn activation
Full SKILL.md body loads when the agent decides the task matches. Commands and agents load on invocation.
< 5,000 tokens per skillOn demand
Reference files, scripts, and assets load only when the skill instructions reference them explicitly.
zero cost if unusedDesigned to coexist,
not replace.
ACS works alongside every existing standard. Add .agents/ to any project without breaking what's there.
AGENTS.md
Keep your AGENTS.md for broad compatibility. ACS adds structured layers on top. They live side by side in the same repo.
SKILL.md (agentskills.io)
ACS skills use the exact same SKILL.md format. Any agentskills.io skill works inside .agents/skills/ with zero changes.
Model Context Protocol
MCP handles runtime tools. ACS handles project configuration. Different layers, fully complementary.
CLAUDE.md
Migrate incrementally. The planned acs compile CLI will generate CLAUDE.md from ACS automatically.
Your project,
agent-ready in minutes.
Create a .agents/ folder. Add main.yaml. Write context. Done — any ACS-compatible agent picks it up automatically.