Which applications does it support?

It is optimized for common AppWorld integrations including Spotify, Venmo, Gmail, Contacts, and Calendar APIs.

What is the primary purpose of the reflect-appworld-failure skill?

It is designed to perform a semantic analysis of why an agent failed a task in AppWorld and turn that failure into a permanent 'lesson' or playbook entry for future use.

What kind of errors can this skill handle?

It specializes in analyzing API misuse, authentication order issues, incorrect method naming conventions, and data structure access errors.

How does it improve future AI performance?

By generating structured JSON bullets with working code examples, it provides high-quality context that helps the agent generator avoid making the same mistake twice.

AppWorld API Pattern Reflector

Name: AppWorld API Pattern Reflector
Author: jmanhype

byjmanhype

•

セキュリティとテスト

Analyzes failed AppWorld tasks to extract specific API usage patterns and generate actionable remediation playbooks.

概要

This skill automates the post-mortem analysis of failed agentic tasks within the AppWorld environment. By examining error logs, failed code snippets, and execution context, it identifies the root cause—such as API misuse, incorrect authentication sequences, or non-existent method calls—and transforms these failures into reusable, high-quality knowledge bullets. It ensures that future agent executions learn from past mistakes by providing concrete code examples and verified patterns for various integrated services like Spotify, Venmo, and Google Workspace.

主な機能

Semantic categorization and tagging of failures for future retrieval
Confidence scoring for learned patterns to ensure knowledge reliability
Automated extraction of specific API method naming conventions
9 GitHub stars
Root cause analysis of API execution failures and logic errors
Generation of actionable playbook bullets with working code examples

ユースケース

Debugging failed agentic workflows in the AppWorld benchmarking environment
Converting raw execution error logs into structured, readable developer documentation
Building a self-improving knowledge base for multi-agent trading and automation systems

概要

主な機能

Semantic categorization and tagging of failures for future retrieval
Confidence scoring for learned patterns to ensure knowledge reliability
Automated extraction of specific API method naming conventions
9 GitHub stars
Root cause analysis of API execution failures and logic errors
Generation of actionable playbook bullets with working code examples

ユースケース

Debugging failed agentic workflows in the AppWorld benchmarking environment
Converting raw execution error logs into structured, readable developer documentation
Building a self-improving knowledge base for multi-agent trading and automation systems