The Agent Skills Directory

[SAFE]: The skill is purely educational and provides standard Python code snippets for building resilient systems. It covers best practices for defining exception hierarchies, implementing retry logic with backoff, and using circuit breakers to prevent cascading failures.
[SAFE]: No indicators of prompt injection, data exfiltration, or obfuscation were found. All code uses standard library modules (time, enum, typing) and does not involve risky operations like subprocess execution, network requests, or sensitive file access.
[SAFE]: The restricted toolset defined in the frontmatter (Read, Grep, Glob) limits the agent's scope to safe read-only operations during skill execution.

error-handling