goal-loop-orchestrator

Execute goals to terminal verdicts with transcript evidence and stall detection.

2|Updated Jul 13, 2025

Search Tags:#stall detection #autonomous workflow #goal execution #turn budgeting #proof of progress

Ready to run

npx skills add https://github.com/krzemienski/shannon --skill goal-loop-orchestrator

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: goal-loop-orchestrator
Source: https://github.com/krzemienski/shannon/tree/main/skills/goal-loop-orchestrator
Command: npx skills add https://github.com/krzemienski/shannon --skill goal-loop-orchestrator

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill ensures goal completion in autonomous workflows, providing proof of progress, detecting stalls, and managing turn budgets.

Core Features & Use Cases

Goal Completion Automation: Execute goals to completion with transcript evidence, stall detection, and turn/token budgeting.
Stall Detection: Identifies when a goal is not making progress and stops execution.
Turn Budgeting: Manages the number of turns a goal can take to prevent resource exhaustion.
Use Case: When you need to run a complex task like 'refactor the codebase to improve performance' without manual intervention, the Skill will run until completion, report stalls, and manage turns.

Quick Start

Run a goal to completion using the command: /goal run /refactor_codebase_to_improve_performance

Frequently Asked Questions about goal-loop-orchestrator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I automate goal execution to completion without manual intervention?▼

Automate goal execution to completion by running tasks until a terminal verdict is reached, using transcript evidence and turn budgeting to manage resources. This prevents indefinite loops during complex autonomous workflows.

What is stall detection in autonomous workflows and how does it work?▼

Stall detection in autonomous workflows identifies when a goal is no longer making meaningful progress and stops execution. This prevents resource exhaustion by halting tasks that are stuck in unproductive loops.

How do I prevent resource exhaustion during long-running autonomous tasks?▼

Prevent resource exhaustion in autonomous tasks by applying turn budgeting, which limits the number of turns or tokens a goal can consume. Execution stops automatically if the budget is exceeded.

How do I run a complex codebase refactor autonomously with proof of progress?▼

Run a complex codebase refactor autonomously by executing the goal with a defined condition. The system provides transcript evidence as proof of progress and stops when the goal is satisfied or stalled.

Do I need a predefined goal condition to use autonomous workflow execution?▼

Yes, autonomous workflow execution requires a well-defined goal condition from a goal-condition-architect and a valid run state. These prerequisites ensure the system can verify completion and detect stalls accurately.

What happens when an autonomous goal cannot be completed?▼

When an autonomous goal cannot be completed, the system stops execution and returns a BOUND_REACHED or STALLED terminal verdict. This outcome includes transcript evidence showing exactly where progress halted.