task-strategy

Classify tool failures and route to safe fallbacks or user escalation.

9|Updated Jun 11, 2026
One-click install
npx skills add https://github.com/llm011/ethan-agent --skill task-strategy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: task-strategy
Source: https://github.com/llm011/ethan-agent/tree/main/ethan/defaults/skills/task-strategy
Command: npx skills add https://github.com/llm011/ethan-agent --skill task-strategy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents endless retry loops when tools fail, operations are denied, or tasks time out by guiding the agent toward appropriate fallback paths or clear user escalation.

Core Features & Use Cases

  • Failure Classification: Distinguishes unavailable services, authorization denials, timeouts, parameter errors, and unknown failures.
  • Fallback Routing: Provides alternatives for browser, shell, delegated coding, and computer-use failures without bypassing permissions.
  • Safety Guardrails: Stops repeated attempts after three failures and prevents using different tools to circumvent the same restriction.
  • Use Case: When a browser session is unavailable, the Skill directs the agent to use an alternative browser tool or report the blocker instead of repeatedly retrying the failed session.

Quick Start

Use the task-strategy skill to choose a safe fallback after a tool call fails or a task cannot be completed.

Frequently Asked Questions about task-strategy

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

FAQPage Schema
How do I stop an AI agent from endlessly retrying failed tool calls?

To stop endless retries on failed tool calls, you need a task strategy that classifies the failure and routes to a safe fallback or escalates the blocker. This halts repeated attempts after three failures and prevents bypassing permissions.

What is the best way to handle browser recovery when a session becomes unavailable?

Handling browser recovery requires routing to an alternative browser tool or reporting the blocker instead of retrying. This fallback strategy prevents repeated attempts on the unavailable session and ensures operations respect existing permissions.

How do I manage task escalation when shell authorization is denied?

Managing task escalation after shell authorization denials involves selecting permission-respecting alternatives or reporting the blocker to the user. This strategy avoids using different tools to circumvent the same restriction and stops repeated failure loops.

Can I use fallback routing for timeouts and unavailable computer-use services?

Yes, fallback routing applies to timeouts and unavailable computer-use services. The strategy requires failure classification to distinguish these issues and provides bounded retries, ensuring the agent selects safe fallback actions instead of looping.

Why does an agent keep trying different tools to bypass a permission restriction?

An agent tries bypassing permission restrictions when lacking a safety guardrail for failure classification. Implementing a task strategy prevents using different tools to circumvent denials, stopping repeated attempts after three failures and ensuring structured blocker reporting.