The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions define operational boundaries and role definitions without any attempts to bypass safety filters or override system-level constraints. The language is purely instructional and focused on architectural design.
[DATA_EXFILTRATION]: No commands for file system access or network transmission were detected. The skill does not access sensitive directories or transmit data to external domains.
[REMOTE_CODE_EXECUTION]: There are no patterns involving remote script execution, package installation, or the use of dynamic execution functions like eval or exec.
[COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or scripts that interact with the operating system.
[CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or private tokens were found in the skill files or metadata.
[EXTERNAL_DOWNLOADS]: The skill does not download external resources or define remote dependencies.
[NO_CODE]: The skill is composed strictly of documentation, guiding principles, and metadata. It lacks any executable scripts or automation logic that could be leveraged for malicious purposes.

worker-architect