The Agent Skills Directory

[SAFE]: No security threats or malicious patterns were detected across the root skill or its dedicated sub-modes. The instructions follow standard agent orchestration patterns intended for software development workflows.
[PROMPT_INJECTION]: The instructions define operational roles (Lead, Executor) and quality standards without attempting to bypass safety guardrails, extract system prompts, or override platform constraints.
[DATA_EXFILTRATION]: No unauthorized network calls or attempts to access sensitive system files, such as environment variables, SSH keys, or cloud credentials, were identified.
[COMMAND_EXECUTION]: Instructions for running build, test, and lint commands in $work-init and $work are consistent with the skill's purpose as a developer tool and do not include hardcoded malicious payloads or unauthorized privilege escalation attempts.
[EXTERNAL_DOWNLOADS]: References to external repositories and package registries are limited to the vendor's own resources (xcaeser) or official platform schemas, representing standard functionality for skill management.

work