The Agent Skills Directory

[SAFE]: The skill is focused on improving code quality and security through static analysis. It includes a recommended configuration file for golangci-lint that enables several security-oriented linters, such as gosec (for detecting SQL injection, hardcoded credentials, and weak crypto), bidichk (for detecting dangerous bidirectional Unicode sequences), and bodyclose/sqlclosecheck (for preventing resource leaks).
[SAFE]: The skill provides clear instructions for using //nolint directives safely, mandating the inclusion of specific linter names and justification comments. This prevents blanket suppression of linting rules and ensures that security-related warnings are not inadvertently silenced.
[SAFE]: Tool access is appropriately scoped to development-related tasks (go, golangci-lint, git) and standard file operations. No evidence of credential harvesting, unauthorized network communication, or persistence mechanisms was found.
[SAFE]: The skill uses the Agent tool for parallelizing code cleanup tasks, which is a legitimate orchestration pattern within the agent's operating environment and does not introduce external risks.

golang-lint