The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool, restricted to go:*, golangci-lint:*, and git:* command patterns. These tools are used for standard development tasks such as module upgrades, running tests (go test), performing vulnerability scans (govulncheck), and applying automated fixes (go fix).
[EXTERNAL_DOWNLOADS]: The skill fetches documentation and changelogs from official Go sources (e.g., go.dev). It also references the Go vulnerability database at vuln.go.dev for security scanning. These are well-known, trusted services within the Go ecosystem.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process project files, including Go source code, go.mod module definitions, and a local .modernize configuration file. While these represent untrusted input sources, the skill mitigates risks by requiring verification of all edits through the official Go compiler, test suite, and linters (golangci-lint) before completion.
Ingestion points: Reads go.mod, .modernize, and all **/*.go files in the workspace.
Capability inventory: Uses Bash for Go tooling, Edit/Write for file modifications, and Agent for parallelized scanning.
Sanitization: Relies on go test, golangci-lint, and the stdversion vet check to validate correctness and safety of generated code changes.
Boundary markers: None explicitly defined in the prompt templates, but the skill scope is strictly limited to modernization patterns.

golang-modernize