- Home
- Skills
- Agents & Orchestration
- Freebies — Simple Subagent Delegator
Works with the AI Agents you already use
Freebies — Simple Subagent Delegator
Decompose complex tasks into specialized subagent workflows with explicit context handshakes and state isolation.
Free
Freebies — Simple Subagent Delegator
Example session with this skill installed
Break down the task of building a secure auth service into a subagent delegation plan with clear handoffs.
- Read your context and instructions
- Compiled the freebies
I have decomposed the authentication service development into a modular subagent workflow. This plan establishes four specialized personas with strict input/output contracts to ensure security and architectural integrity.
Subagent Delegation Orchestration Plan
| Phase | Subagent Persona | Input Contract | Output Deliverable | Verification Gate |
|---|---|---|---|---|
| 1 | Security Architect | Feature Requirements | auth-spec.md (Threat Model) | OWASP Top 10 Checklist |
| 2 | Database Designer | auth-spec.md | schema.sql (Bcrypt/Salted) | Normalization Review |
| 3 | Backend Developer | schema.sql + Spec | auth-service.ts | Unit Test Coverage > 90% |
| 4 | Security Reviewer | auth-service.ts | Audit Report | Pentest Simulation |
Message Passing Schema (JSON Handoff)
{
"trace_id": "auth-svc-001",
"source_agent": "Security-Architect",
"target_agent": "Database-Designer",
"payload": {
"encryption_standard": "AES-256-GCM",
"hashing_rounds": 12,
"session_strategy": "JWT-Stateless"
},
"verification_status": "PASSED"
}
Next steps
- Review the
auth-spec.mdrequirements for JWT expiration policies. - Initialize the Database Designer subagent with the
schema.sqltemplate. - Run the automated evaluation suite in
evals/eval-suite.jsonto validate the handoff logic.
Connects securely to your tools. The creator never sees your data.
About this skill
Monolithic prompts often fail when tasks become too complex for a single context window. This skill provides a deterministic framework for decomposing high-level goals into modular subagent workflows with strict state isolation and contract-based handoffs.
What it does
- Task Decomposition breaks down complex user goals into independent, sequentially dependent subtasks.
- Persona Assignment routes subtasks to specialized roles like Researcher, Architect, or Reviewer.
- Context Handshaking defines explicit Markdown or JSON contracts for data passing between agents.
- Verification Checkpoints establishes quality gates before intermediate outputs advance to the next stage.
- State Isolation ensures subtasks remain modular and free from global context pollution.
How it works
- Analyze the high-level request to identify logical boundaries for sub-orchestration.
- Define the specialized subagent personas required for the specific technical domain.
- Map the sequential flow and data contracts required for each agent handoff.
- Output a structured orchestration diagram and message-passing schema for implementation.
Frameworks & tools
Built for AAIF agentskills.io Open Standard. Compatible with Claude Code, Cursor, Windsurf, Gemini CLI, and OpenAI Codex. Supports project-scoped runners like npx, uvx, and pipx.
Why this beats prompting it yourself
Manual delegation often leads to "context drift" where agents lose track of the original goal. This skill enforces strict handshake contracts and verification steps that prevent logic errors during complex, multi-stage agentic workflows.
Use cases
- Designing a multi-agent swarm to handle automated PR reviews and refactoring.
- Orchestrating a research-to-code pipeline with distinct architect and implementer roles.
- Implementing MCP server workflows that require isolated state between tool executions.
Known limitations
Restricted to the active project workspace. Does not support global system package installations or root-level commands.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Workspace-confined to ./. No support for parent directory access, root/admin commands, or global package installations.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed today
- Free with an Agensi account
Linux only
- Software
- Zero global installationorelevated privileges required.
File System Read/Write (strictly scoped to current project workspace `./`).
Permissions
Target Use Cases: - Design a subagent task delegation workflow. - Break down a complex task for multiple AI agents. - Create handoff contracts between AI subagents.