When does the skill stop trying to recover?

The skill escalates to human intervention when it encounters 'Unrecoverable' errors (like Out of Memory or Disk Full) or when the defined retry limits for a specific error have been exceeded.

How does Error Recoverer decide which strategy to use?

It uses an internal classifier to determine the error category and severity, then queries causal memory for similar past issues to select strategies like 'Retry', 'Rollback', or 'Fix and Retry'.

What types of errors can Error Recoverer handle?

The skill handles a wide range of errors including transient network timeouts, resource permission issues, syntax/parsing errors, logic test failures, and environment dependency mismatches.

Can this skill automatically fix broken code?

Yes, for syntax and specific logic errors, Error Recoverer can trigger a 'Fix and Retry' workflow that attempts to debug and resolve the code issue before continuing.

Error Recoverer

Name: Error Recoverer
Author: adaptationio

byadaptationio

0•

安全与测试

Automates error detection, classification, and recovery to maintain continuous and resilient autonomous coding sessions.

The Error Recoverer skill provides Claude with a sophisticated framework for identifying and resolving issues that occur during autonomous development tasks. By classifying errors into specific categories such as transient, resource, syntax, or logic, it intelligently selects the optimal recovery strategy—ranging from exponential backoffs and state rollbacks to automated code fixes. This skill is essential for maintaining session continuity, reducing manual intervention, and ensuring that coding agents can autonomously navigate through failures by leveraging causal memory and integrated checkpoint management.

主要功能

01Multi-tier recovery strategies including exponential backoff, state rollback, and fix-and-retry

02Detailed escalation protocols that generate comprehensive reports for human intervention

03Intelligent error classification across transient, resource, syntax, logic, and environment categories

04Automated checkpoint management for safe state restoration after unrecoverable failures

05Causal memory integration to query and store successful error-to-solution chains

060 GitHub stars

使用场景

01Managing transient network timeouts and rate limits during complex API integrations

02Automatically correcting syntax or JSON parsing errors generated during autonomous code creation

03Restoring stable project states when logic changes lead to regression test failures

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add adaptationio/skrillz error-recoverer

For use in Claude.ai and ChatGPT

Download Skill