Aws Agentic Ai
Deploy Claude Agent SDK, OpenClaw, or Strands-style agents with S3-backed shared SKILL.md and config so capability updates do not require redeploys.
Overview
aws-agentic-ai is an agent skill most often used in Build (also Operate infra) that deploys filesystem-based agent frameworks with S3-mounted shared skills and config on AWS AgentCore.
Install
npx skills add https://github.com/zxkane/aws-skills --skill aws-agentic-aiWhat is this skill?
- Maps Claude Agent SDK, OpenClaw, and Strands Agents config discovery to a shared S3 Files NFS layout
- Update SKILL.md or CLAUDE.md in S3 and agent instances pick up new capabilities without redeployment
- Documents AgentCore + S3 Files architecture for multi-instance agent fleets
- Tables tie each framework to key config files and discovery mechanisms (cwd, gateway dir, Python modules)
- Applies to Runtime, S3 Files, and filesystem-based agent frameworks on AWS
- Compares three agent frameworks (Claude Agent SDK, OpenClaw, Strands Agents) with config discovery tables
Adoption & trust: 1 installs on skills.sh; 306 GitHub stars; 2/3 security scanners passed (skills.sh audits); trending (+100% hot-view momentum).
What problem does it solve?
Every agent instance has its own copy of skills and CLAUDE.md, so fixing a SKILL.md means redeploying the fleet and configs drift out of sync.
Who is it for?
Indie builders running multiple Claude or OpenClaw agents on AWS who want hot-updatable skills and shared project settings at scale.
Skip if: Single-local-agent hobby setups with no AWS footprint, or teams that only need generic Lambda snippets without S3 Files NFS mounting.
When should I use this skill?
Deploying or scaling filesystem-based agent frameworks on AWS where CLAUDE.md, SKILL.md, OpenClaw, or Strands tooling must stay consistent across instances via S3 Files.
What do I get? / Deliverables
Agents read a single S3-backed filesystem for CLAUDE.md and skill trees, so configuration edits propagate to all instances on the next file access without redeployment.
- Architecture for S3-backed shared agent configuration trees
- Mapped layout of framework config files to bucket paths
- Operational workflow to update skills without redeploying agents
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
You adopt this while assembling agent runtimes and shared skill corpora—the core Build/agent-tooling moment—before day-two tuning moves to Operate. Focus is filesystem-discovered agent configuration (CLAUDE.md, .claude/skills, OpenClaw trees) mounted for multiple instances, not generic CRUD APIs.
Where it fits
Lay out `.claude/skills` and CLAUDE.md on S3 Files before launching your first AgentCore fleet.
Wire OpenClaw gateway working directories to the same bucket path as Claude SDK project settings.
Patch a SKILL.md in S3 during an incident response so all agents pick up a new tool policy without rolling pods.
How it compares
AWS deployment pattern for shared agent config—not a local skills.sh installer or a non-AWS container recipe.
Common Questions / FAQ
Who is aws-agentic-ai for?
Developers using Claude Code-family agents or OpenClaw on AWS who need centralized SKILL.md and CLAUDE.md distribution across AgentCore runtimes.
When should I use aws-agentic-ai?
In Build/agent-tooling when designing multi-instance agent hosting; in Build/integrations when connecting S3 Files to your runtime; and in Operate/infra when updating live skills without redeploying containers.
Is aws-agentic-ai safe to install?
It describes cloud infrastructure patterns that imply S3 and runtime access—review IAM boundaries and the Security Audits panel on this page before applying in production accounts.
SKILL.md
READMESKILL.md - Aws Agentic Ai
# Deploying Filesystem-Based Agent Frameworks with S3 Files and AgentCore **Applies to**: Runtime, S3 Files, Claude Agent SDK, OpenClaw, Strands Agents ## Core Insight Modern agent frameworks discover capabilities by reading configuration files from the working directory at startup: | Framework | Key Config Files | Discovery Mechanism | |-----------|-----------------|---------------------| | **Claude Agent SDK** | `CLAUDE.md`, `.claude/skills/*/SKILL.md`, `.claude/commands/*.md`, `.claude/output-styles/*.md` | `cwd` + `setting_sources=["project"]` | | **OpenClaw** | `.openclaw/`, `.agents/`, `skills/`, `.codex`, `.env` | Gateway working directory | | **Strands Agents** | Agent code, `requirements.txt`, tool definitions | Python module loading | Modify a `SKILL.md` → agent gains new capabilities. Update `CLAUDE.md` → agent follows new guidelines. No redeployment. **S3 Files** mounts an S3 bucket as a shared NFS filesystem. All agent instances see identical configuration files. Update once in S3 → every instance picks it up on the next file access. ## Architecture ``` ┌────────────────────────────────────────────────────┐ │ S3 Bucket (agent configuration source of truth) │ │ │ │ ├── CLAUDE.md │ │ ├── .claude/skills/*/SKILL.md │ │ ├── .claude/commands/*.md │ │ ├── .openclaw/skills/ │ │ └── shared-knowledge/ │ └─────────────┬───────────────────────────────────────┘ │ S3 Files NFS (auto bidirectional sync) ▼ ┌─────────────────┐ ┌──────────────────┐ ┌──────────────┐ │ EC2 / EKS / ECS │ │ EC2 / EKS / ECS │ │ Lambda │ │ /mnt/s3-config/ │ │ /mnt/s3-config/ │ │ /mnt/s3/ │ │ (Claude Agent) │ │ (OpenClaw) │ │ (Strands) │ └─────────────────┘ └──────────────────┘ └──────────────┘ ``` **AgentCore Runtime** microVMs cannot mount S3 Files directly. For AgentCore deployments, pull shared config from S3 via API at session start, or use Session Storage for per-session file persistence. ## Claude Agent SDK Deployment ### With S3 Files (EC2/EKS) Point `cwd` to the S3 Files mount. The SDK reads `CLAUDE.md`, skills, commands, and output styles transparently: ```python from claude_agent_sdk import query, ClaudeAgentOptions S3_MOUNT = "/mnt/s3-config" # S3 Files mount point async def handle_request(prompt: str, session_id: str = None): options = ClaudeAgentOptions( cwd=S3_MOUNT, setting_sources=["project"], allowed_tools=["Skill", "Read", "Write", "Edit", "Bash", "Glob", "Grep"], ) if session_id: options.resume = session_id async for message in query(prompt=prompt, options=options): if hasattr(message, "result"): yield message.result ``` Enable Bedrock model access (no Anthropic API key required): ```bash export CLAUDE_CODE_USE_BEDROCK=1 # AWS credentials from instance role ``` ### Zero-Downtime Skill Updates Upload a new skill to S3 — all running instances discover it on next invocation without restart: ```bash aws s3 cp ./api-testing/SKILL.md \ s3://agent-configs/.claude/skills/api-testing/SKILL.md ``` ### On AgentCore Runtime (S3 API Sync) Pull shared config from S3 to Session Storage at session start: ```python import boto3, os from bedrock_agentcore.runtime import BedrockAgentCoreApp from claude_agent_sdk import query, ClaudeAgentOptions import asyncio app = BedrockAgentCoreApp() s3 = boto3.client("s3") WORKSPACE = "/mnt/workspace" def sync_config(): """Pull shared agent config from S3 to local workspace.""" for prefix in ["CLAUDE.md", ".claude/skills/", ".claude/commands/"]: if "/" not in prefix: obj = s3.get_object(Bucket="agent-configs", Key=prefix) path = f"{WORKSPACE}/{prefix}" with open(path, "w") as f: f.write(obj["Body"]