The Agent Skills Directory

[Instructional Intent]: The skill is designed to guide an AI agent through the process of auditing code for breaking changes in external integration surfaces. This is a standard software engineering task.
[Absence of Risky Patterns]: The skill contains no executable code, shell commands, network requests, or external dependencies. It relies entirely on the agent's internal reasoning and standard file-reading capabilities.
[Indirect Prompt Injection Surface]: As a code analysis skill, the agent will ingest external code which could theoretically contain malicious instructions. This is a standard consideration for analysis tasks, and the skill does not grant the agent any additional capabilities that could be exploited via such an injection.

code-breaking-changes