golang-code-style

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided Go source code for review and refactoring, which constitutes a surface for indirect prompt injection.
  • Ingestion points: Go source files read using the Read tool.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded content are provided.
  • Capability inventory: The skill uses Bash (restricted to Go tools), Edit, Write, and the Agent tool for sub-agent delegation.
  • Sanitization: No specific sanitization or filtering of input code is defined.
  • [EXTERNAL_DOWNLOADS]: The skill recommends the use of the github.com/samber/lo library and references the author's GitHub repository for project homepages. These are neutral references to the author's own development resources.
  • [COMMAND_EXECUTION]: The skill configuration scopes shell command execution to specific, expected tools for Go development (go, golangci-lint, git). This restriction adheres to the principle of least privilege.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 04:50 AM
Security Audit — agent-trust-hub — golang-code-style