The Agent Skills Directory

[PROMPT_INJECTION]: The instructions focus on architectural enforcement and development workflows. No override markers, bypass attempts, or safety filter circumvention patterns were detected.
[DATA_EXFILTRATION]: There are no network-bound operations or access to sensitive local file paths like SSH keys or credentials. The skill operates purely on the project's Go source files.
[CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or private tokens are present. Code templates use standard placeholders and local database connection examples.
[REMOTE_CODE_EXECUTION]: No remote scripts are downloaded or executed. The skill uses internal references and standard Go dependency management (go.mod).
[COMMAND_EXECUTION]: The workflow spawns specific sub-agents (e.g., 'fuse-ai-pilot:sniper') via an internal 'TeamCreate' mechanism. This is a vendor-native orchestration pattern and does not allow for arbitrary shell command execution.
[EXTERNAL_DOWNLOADS]: No external resources are fetched from untrusted domains. All templates and references are contained within the skill's local 'references/' directory.
[SAFE]: Detailed analysis of obfuscation, privilege escalation, persistence, and dynamic execution categories revealed no suspicious activities. The skill follows best practices for educational and development-focused AI extensions.

solid-go