The Agent Skills Directory

[SAFE]: The skill defines a set of internal guidelines and 'tests' for code quality (Deletion Test, Two-Adapters Rule, Inline Test, etc.) intended to guide an AI agent's decision-making process. These are purely instructional and do not involve executable code.
[SAFE]: The skill mentions managing project-specific documentation in runbooks/ and updating files like CLAUDE.md or AGENTS.md. These actions are within the expected scope of a software engineering assistant and do not constitute a malicious persistence mechanism.
[SAFE]: References to other modules (e.g., desplega:delegate-work, desplega:feedback) appear to be references to other skills within the same trusted environment and do not involve untrusted external dependencies.
[SAFE]: No obfuscation, prompt injection, or data exfiltration patterns were detected in the instructions or metadata.

engineering-standards