golang-how-to

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the gopls language server using go install golang.org/x/tools/gopls@latest. This is a trusted, official repository for Go development tools.
  • [COMMAND_EXECUTION]: The skill leverages go, git, and gopls binaries through bash commands. These tools are explicitly restricted in the allowed-tools configuration to prevent unauthorized command execution by limiting access to specific binaries.
  • [PROMPT_INJECTION]: The skill reads project configuration files (like CLAUDE.md and AGENTS.md) and scans the codebase for library imports to suggest relevant Go skills. This exposes an indirect prompt injection surface where untrusted project data enters the agent's context. Evidence: 1. Ingestion points: CLAUDE.md, AGENTS.md, go.mod, and codebase (via Grep). 2. Boundary markers: Absent. 3. Capability inventory: Read, Write, Edit, and restricted Bash tools. 4. Sanitization: None described. This surface is inherent to the tool's primary purpose and is considered low risk in this context.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 02:20 AM
Security Audit — agent-trust-hub — golang-how-to