The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to review user-provided project files such as go.mod to identify banned dependencies. This creates a theoretical surface where malicious instructions embedded in project files (such as comments) could attempt to influence the agent's behavior during the review process.\n
Ingestion points: SKILL.md (Process step 1: Check go.mod against the banned libraries list)\n
Boundary markers: Absent\n
Capability inventory: references/rules.md (Lists commands like go build, go test, and go install for CI/CD configuration)\n
Sanitization: Absent\n- [EXTERNAL_DOWNLOADS]: The skill provides instructions for fetching the latest Go version from golang.org and installing security auditing tools like gosec, govulncheck, and gitleaks. These references target official repositories and well-known services for the Go ecosystem.\n- [COMMAND_EXECUTION]: The documentation includes common Go development commands (e.g., go build, go test, sqlc generate, templ generate) for use in development and CI/CD pipelines. These commands are standard for Go development and are presented as reference material.\n- [DATA_EXPOSURE]: The skill actively promotes security best practices, such as using .env files for secrets, avoiding the commitment of binary files to version control, and implementing secret-leak detection in CI. No hardcoded credentials or unauthorized data access patterns were found.

how-to-golang