smithery.ai

Delegating to AWS Agent

Recognize AWS queries and delegate to specialized sub-agent to avoid context pollution

Installation

$ npx skills add https://smithery.ai

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from smithery.ai · top by installs.

npx skills add https://smithery.ai

Browse all from smithery.ai

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Skill metadata

Parsed from SKILL.md frontmatter.

Version1.0.0

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,926 B
  • docs SUMMARY.md 117 B

History

  1. First recorded snapshot · 1 installs

SKILL.md

Delegating to AWS Agent

Core Principle

Never handle AWS operations directly. Always delegate to a specialized sub-agent to keep your context clean and costs low.

Recognition Patterns

Delegate when user says:

  • "list ec2 instances"
  • "show s3 buckets"
  • "what access keys does user X have?"
  • "deploy SERVICE to production"
  • "show running instances"
  • "what's in bucket X?"
  • Any mention of: aws, ec2, s3, iam, ecs, lambda, cloudwatch, rds, bucket, instance

How to Delegate

Use the Task tool with a specialized prompt:

Task(
  subagent_type: "general-purpose",
  description: "Query AWS infrastructure",
  prompt: "<full agent instructions from AGENT-INSTRUCTIONS.md>"
)

Agent Prompt Template

When delegating, include:

  1. The complete agent instructions (see AGENT-INSTRUCTIONS.md)
  2. The user's specific request
  3. Clear output format requirements

Example:

You are an AWS infrastructure specialist. Your job is to query AWS using shell wrappers and return clean results.

<AGENT INSTRUCTIONS HERE>

USER REQUEST: List all running EC2 instances

Return a clean summary with:
- Instance IDs
- Names
- States
- Instance types
- IP addresses

After Agent Returns

  1. Present results cleanly to user
  2. Offer follow-up if relevant (e.g., "Would you like details on instance i-abc123?")
  3. Don't expose mechanics (aws commands, authentication, profiles) to user

Benefits

  • ✅ Main context stays clean
  • ✅ Cheaper queries (sub-agent uses less expensive model)
  • ✅ Specialized knowledge isolated
  • ✅ Scalable pattern for other services

Example Flow

User: "list running ec2 instances"

Main Assistant: [Recognizes AWS query]
              → Invokes Task tool with agent instructions
              → Agent runs cloud-ec2-list --state running
              → Agent returns formatted results

Main Assistant: "Found 22 running instances:
                - i-0abc123 (production-web) [running] t3.medium
                - i-0def456 (staging-api) [running] t3.small
                ..."

Red Flags

DON'T:

  • ❌ Try to run cloud-* scripts yourself
  • ❌ Construct aws CLI commands in main session
  • ❌ Load detailed AWS API knowledge
  • ❌ Handle authentication/profiles directly

DO:

  • ✅ Immediately delegate on AWS keywords
  • ✅ Trust the sub-agent's results
  • ✅ Present clean summaries to user

Version History

  • 1.0.0 (2025-10-14): Initial delegation skill created to reduce context pollution