The Agent Skills Directory

[NO_CODE]: The skill consists entirely of Markdown reference files and YAML metadata. It does not include any executable scripts, binaries, or active logic.
[COMMAND_EXECUTION]: The skill provides reference material for powerful Go CLI commands capable of executing arbitrary code or shell processes. Key examples include:
go generate: Runs commands described by //go:generate directives within source files (documented in references/cli.md).
go test and go run: Compile and execute Go code, which can include arbitrary logic defined in tests or main packages (documented in references/test.md and references/cli.md).
go tool: Provides access to low-level toolchain components like the compiler and linker (documented in references/tool.md).
[PROMPT_INJECTION]: The skill facilitates an attack surface for indirect prompt injection by providing instructions for processing untrusted external data (Go projects) using tools with execution capabilities.
Ingestion points: The agent ingests files including Go source code (.go), module files (go.mod), and workspace files (go.work), all of which can be controlled by an attacker in an untrusted repo.
Boundary markers: Absent. The skill does not instruct the agent to use delimiters or ignore embedded instructions when reading project files.
Capability inventory: Subprocess execution via go generate, go run, and go test as described across the reference files.
Sanitization: Absent. There are no instructions for validating, escaping, or filtering the content of directives or source files before invoking CLI tools that execute them.

go