- Home
- Skills
- Workflow Automation
- Vorchestrator: Multi-Agent Quota Balancer
Vorchestrator: Multi-Agent Quota Balancer
Token-efficient multi-agent orchestrator that dynamically tiers subagents across Claude and Gemini quota pools while applying dynamic MCP tool schema pruning.
$19
Works with the AI Agents you already use
Vorchestrator: Multi-Agent Quota Balancer
Example session with this skill installed
Run an autonomous migration of 14 unit test suites to pytest, balancing quota between Gemini and Claude pools.
- Read your context and instructions
- Compiled the vorchestrator
$ # 1. Launch a complex refactoring task with Vorchestrator enabled:
$ vorchestrator run --task "Refactor auth middleware and update 14 test suites"
[VORCHESTRATOR] Task received: 15 subtasks detected. Activating dual-quota balance engine.
$ # 2. Model Tiering Assessment & Subagent Delegation:
[PLAN] Phase 1: Codebase Indexing & Symbol Map
-> Routing to Tier: 'flash_lite' (Gemini Quota Pool) - 0.2s latency, lowest cost
-> MCP Pruning: Disabled 18 irrelevant database and browser tools (saved 14,200 schema tokens)
-> Status: COMPLETED in 3.4s
[PLAN] Phase 2: Refactoring Auth Token Handlers
-> Complexity High: Routing to Tier: 'pro' (Claude Quota Pool)
-> State Handoff: Compact TOON summary passed (no bloated chat history)
-> Status: COMPLETED in 12.1s (4 files modified)
[PLAN] Phase 3: Parallel Test Execution across 14 Suites
-> Fan-out: 4 parallel worker agents on 'flash' tier
-> Quota Check: Claude pool at 42% threshold; balancing tests to Gemini pool
-> Status: 14/14 test suites PASSED
$ # 3. Final Summary & Metrics:
[VORCHESTRATOR] Orchestration finished cleanly.
[METRICS] Total Tokens Saved: 89,450 | Rate Limit Incidents: 0 | Execution Time: 48s
Connects securely to your tools. The creator never sees your data.
About this skill
Agensi Turnkey Convenience Package ($19): While the core codebase remains open-source MIT on GitHub, this official Agensi package delivers turnkey multi-agent orchestration, verified
install.shautomation, curatedSKILL.mdagent rules, dynamic MCP tool schema pruning, ongoing model compatibility updates, and priority support for autonomous dual-provider quota balancing.
Eliminate Rate Limits with Dual-Provider Agent Orchestration
Running large, multi-turn coding sessions on a single LLM provider frequently causes hard rate limits, token quota exhaustion, and escalating API bills.
Vorchestrator is an autonomous multi-agent task orchestrator designed to balance workloads across separate Claude and Gemini rate limit pools. It applies model tiering (flash_lite for read/index lookups, flash for edits, and pro/Claude for high-reasoning architecture) and dynamically prunes unused MCP tool schemas to keep parent context windows lean.
Core Capabilities:
- Dual Quota Pool Balancing: Seamlessly shifts subagent invocations between Claude and Gemini to prevent throttling.
- Dynamic Schema Pruning: Filters large MCP JSON schemas before agent turns, eliminating thousands of wasted overhead tokens.
- Coordinator Successor Pattern: Spawns clean successor coordinators when turns spike, preventing context degradation.
See It in Action (Demo)
$ # 1. Launch a complex refactoring task with Vorchestrator enabled:
$ vorchestrator run --task "Refactor auth middleware and update 14 test suites"
[VORCHESTRATOR] Task received: 15 subtasks detected. Activating dual-quota balance engine.
$ # 2. Model Tiering Assessment & Subagent Delegation:
[PLAN] Phase 1: Codebase Indexing & Symbol Map
-> Routing to Tier: 'flash_lite' (Gemini Quota Pool) - 0.2s latency, lowest cost
-> MCP Pruning: Disabled 18 irrelevant database and browser tools (saved 14,200 schema tokens)
-> Status: COMPLETED in 3.4s
[PLAN] Phase 2: Refactoring Auth Token Handlers
-> Complexity High: Routing to Tier: 'pro' (Claude Quota Pool)
-> State Handoff: Compact TOON summary passed (no bloated chat history)
-> Status: COMPLETED in 12.1s (4 files modified)
[PLAN] Phase 3: Parallel Test Execution across 14 Suites
-> Fan-out: 4 parallel worker agents on 'flash' tier
-> Quota Check: Claude pool at 42% threshold; balancing tests to Gemini pool
-> Status: 14/14 test suites PASSED
$ # 3. Final Summary & Metrics:
[VORCHESTRATOR] Orchestration finished cleanly.
[METRICS] Total Tokens Saved: 89,450 | Rate Limit Incidents: 0 | Execution Time: 48s
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Workflow Automation.
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
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed today
- Software
- Bash 4+Antigravity CLI / Claude Code
Permissions
File Scopes
Orchestrates local agent sessions across configured model providers.