The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of instructions attempting to bypass safety filters, extract system prompts, or override agent behavior was found.
[DATA_EXFILTRATION]: No evidence of hardcoded credentials, sensitive file path access, or unauthorized network operations. References to database connection strings are confined to standard environment variable usage in testing examples.
[REMOTE_CODE_EXECUTION]: The skill does not perform remote downloads or execute arbitrary code from untrusted sources. Referenced tools like golang-migrate and sqlmock are industry-standard development utilities.
[COMMAND_EXECUTION]: The skill utilizes legitimate development tools (Go CLI, golangci-lint, git) as defined in its allowed-tools configuration to assist in code generation and auditing.
[SAFE]: The skill implements strong security boundaries by instructing the agent to use allowlists for dynamic column names and strictly forbidding SQL string concatenation. It also includes measures to prevent resource exhaustion, such as mandatory connection pool configuration and closing database rows.

golang-database