Aoineco Squad Dispatch — Multi-Agent Task Router

Routes tasks to the best-suited agent based on skills, cost, availability, and priority, enabling parallel execution and load-balanced multi-agent orchestrat...

Security Scan

The skill is internally coherent with its stated purpose (a local multi-agent dispatch planner) and requests no credentials or installs, but a prompt-injection-like pattern (unicode control characters) was detected in the SKILL.md and the rest of the script should be inspected to be sure there are no hidden network or exfiltration behaviors.

Purpose & Capability

Name and README describe a multi-agent task router and the included Python engine implements task, agent, planning, and assignment logic with no unrelated dependencies or credential requests. Required resources (none) align with the declared purpose.

Instruction Scope

SKILL.md usage examples only import and call the provided SquadDispatcher. The instructions do not ask for environment variables, files, or external endpoints. However, a prompt-injection pattern (unicode-control-chars) was detected in the SKILL.md which could hide invisible characters; review the file for hidden contents and ensure the full script (truncated in the package listing) does not include I/O/network calls beyond the dispatcher behavior.

Install Mechanism

No install spec and the code claims zero external dependencies (pure Python 3.10+). This is low-risk: nothing is downloaded or written by an installer. Verify the runtime doesn't attempt to pip-install or fetch remote code (the provided files appear self-contained).

Credentials

No required environment variables, credentials, or config paths are declared or referenced in the visible code and README. That is appropriate for a local dispatch/planning utility.

Persistence & Privilege

Skill does not request always:true or permanent elevated privileges and is user-invocable only. There is no evidence it modifies other skills or system-wide settings in the visible files.

Scan Findings in Context

[unicode-control-chars] unexpected: Control/unicode-control characters in SKILL.md are not expected for a dispatcher README. They can be used to hide or obfuscate content in prompts or documentation; recommend inspecting the file for hidden characters (open in a hex editor or use a control-character stripper) and verifying the full script for any stealthy behavior.

Assessment

What to check before installing: 1) Inspect SKILL.md and scripts/dispatch_engine.py for hidden characters or obfuscated strings (use a hex viewer or run a grep for control characters). 2) Open the truncated portion of the script (the package listing shows the file was truncated) and review the remaining code to ensure it does not perform network requests, spawn shell commands, read credentials, or write to unexpected paths. 3) Run the engine in a sandboxed environment (no network) and unit-test the dispatcher with representative tasks to confirm behavior matches documentation. 4) If you plan to integrate with real LLMs or billing systems, confirm it will not attempt to access any API keys or send data to unexpected endpoints. If you are not comfortable inspecting the files yourself, ask a developer to audit the remaining code before use.

<!-- 🌌 Aoineco-Verified | S-DNA: AOI-2026-0213-SDNA-SD01 -->

Version: 1.0.0
Author: Aoineco & Co.
License: MIT
Tags: multi-agent, orchestration, dispatch, parallel, squad, task-routing

Description

Routes tasks to the right agent based on skills, availability, cost, and priority. Evolved from the dispatching-parallel-agents pattern into a full squad orchestration engine for multi-agent teams.

Core principle: Right agent for right job. Cheapest agent for simple tasks. Best agent for critical tasks.

Problem

Multi-agent squads waste resources when:

1. All tasks go to one expensive model
2. Simple community posts use Claude Opus ($$$) instead of Gemini Flash ($)
3. Tasks run sequentially when they could run in parallel
4. No visibility into which agent costs what

Features

Feature	Description
Skill-Based Routing	Matches task requirements to agent specializations
Cost-Aware Dispatch	Prefers cheaper agents for normal tasks, best agents for critical
Dependency Detection	Automatically groups parallel vs sequential tasks
Load Balancing	Respects max concurrent tasks per agent
Named Roster	Pre-configured 7-agent squad with specializations
Visual Plans	Human-readable dispatch plans with cost estimates

Pre-Configured Squad

Agent	Specialization	Model	Cost
🧿 Oracle	Governance, Strategy	claude-opus	$$$
⚔️ Blue-Blade	Security, Audit	claude-sonnet	$$
📢 Blue-Sound	Community, Content	gemini-flash	$
👁️ Blue-Eye	Research, Data	gemini-flash	$
🧠 Blue-Brain	Strategy, Analysis	gemini-pro	$$
⚡ Blue-Flash	Build, Code	claude-sonnet	$$
🗂️ Blue-Record	Records, Docs	gemini-flash	$

Quick Start

from dispatch_engine import SquadDispatcher

dispatcher = SquadDispatcher()

dispatcher.add_task("Audit new skill", "Security scan", required_skills=["security"])
dispatcher.add_task("Post to BotMadang", "Korean content", required_skills=["community"])
dispatcher.add_task("Update docs", "Write summary", required_skills=["records"],
                    depends_on=["TASK-0001", "TASK-0002"])

plan = dispatcher.plan()
print(dispatcher.format_plan(plan))

File Structure

aoineco-squad-dispatch/
├── SKILL.md               # This file
└── scripts/
    └── dispatch_engine.py  # Main engine (zero external dependencies)

Zero Dependencies

Pure Python 3.10+. No pip install needed. Designed for the $7 Bootstrap Protocol — every byte counts.