agentleFS
ControllableModelControllableModel/SWE-Skills-Bench/skills/cobra-cli-development/SKILL.md

cobra-cli-development

Skill8 starsChanged 26 days agoMITView on GitHub

Golang CLI command tree library using spf13/cobra — cobra.Command, RunE vs Run, PersistentPreRunE hook chain, Args validators (NoArgs, ExactArgs, MatchAll, custom), persistent vs local flags, command groups, ValidArgsFunction, RegisterFlagCompletionFunc, ShellCompDirective, usage/help template customization, man-page and markdown doc generation, and testing with SetArgs/SetOut/SetErr. Apply when using or adopting spf13/cobra, or when the codebase imports `github.com/spf13/cobra`. For configuration layering alongside cobra, see the `samber/cc-skills-golang@golang-spf13-viper` skill. For general CLI architecture (project layout, exit codes, signal handling, I/O patterns), see `samber/cc-skills-golang@golang-cli`.

  • Installs packages

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 1 is distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Every cobra CLI has a root command plus zero or more subcommands registered with AddCommand.

    line 62
  2. Very specific:

    Use AddGroup to label subcommands in help output — register groups before the AddCommand calls that reference them.

    line 73
  3. Very specific:

    Always use *E variants — the non-E forms cannot return errors.

    line 83
  4. Very specific:

    Test commands by executing them programmatically.

    line 124
  5. Distinctive:

    If you encounter a bug or unexpected behavior in spf13/cobra, open an issue at <https://github.com/spf13/cobra/issues>.

    line 170

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Persona: You are a Go CLI engineer building command trees that feel native to the Unix shell. You design the user-facing surface first, then wire behavior into the right hook.

Modes:

  • Build — creating a new CLI from scratch: follow command tree setup, hook wiring, and flag sections sequentially.
  • Extend — adding subcommands, flags, or completions to an existing CLI: read the current command tree first, then apply changes consistent with the existing structure.
  • Review — auditing an existing CLI: check the Common Mistakes table, verify RunE usage, OutOrStdout(), hook chain ordering, and args validation.

Using spf13/cobra for CLI command trees in Go

Cobra is the de facto standard for Go CLI applications. It provides the command/subcommand tree, flag parsing (via pflag), args validation, shell completion generation, and documentation generation. It does not handle configuration layering — that's viper's job.

Official Resources:

This skill is not exhaustive. Please refer to library documentation and code examples for more information. Context7 can help as a discoverability platform.

go get github.com/spf13/cobra@latest

Cobra vs. viper

These libraries do fundamentally different things and can be used independently.

Concerncobraviper
OwnsCommand tree, flags, arg validation, completionsConfiguration value resolution
User-facing?Yes — subcommands, flags, help textNo — purely a key-value resolver
Without the other?Yes — a CLI with flags only needs cobraYes — a daemon reading YAML + env needs only viper
Integration seamHands pflag.Flag to viper via BindPFlagTreats the cobra flag as the highest-precedence layer

Use cobra alone when your binary takes flags and args but needs no config file or env resolution. Use viper alone when you have a long-running service reading config from YAML + env with no CLI subcommands. Use both when you need both — bind at PersistentPreRunE on the root command.

→ See samber/cc-skills-golang@golang-spf13-viper for the viper side of this integration.

Command tree

Every cobra CLI has a root command plus zero or more subcommands registered with AddCommand. The root command name is the binary name.

var rootCmd = &cobra.Command{
    Use:          "myapp",
    Short:        "One-line summary",
    SilenceUsage: true,  // ✓ prevents usage wall on every error
    SilenceErrors: true, // ✓ lets you control error output format
}

Use AddGroup to label subcommands in help output — register groups before the AddCommand calls that reference them; cobra does not retroactively assign groups.

The Run* family

Cobra commands have five run hooks executed in order:

PersistentPreRunE → PreRunE → RunE → PostRunE → PersistentPostRunE

Always use *E variants — the non-E forms cannot return errors. Key rules:

  • PersistentPreRunE on the root runs before every subcommand — use it for config init and auth checks.
  • A child PersistentPreRunE replaces the parent's entirely — call the parent explicitly if you need both.
  • PostRunE runs only if RunE succeeded.

For the full lifecycle and inheritance rules, see commands-and-args.md.

Args validators

Cobra validates positional arguments before RunE runs. Never write len(args) checks inside RunE — that bypasses cobra's standard error messages and arg count tracking.

Built-ins: NoArgs, ExactArgs(n), MinimumNArgs(n), MaximumNArgs(n), RangeArgs(min,max), OnlyValidArgs, ExactValidArgs(n). Compose with MatchAll(v1, v2). Custom validator: func(cmd *cobra.Command, args []string) error.

For the full validator set with examples and MatchAll patterns, see commands-and-args.md.

Flags primer

Cobra delegates flag parsing to pflag. Persistent flags (PersistentFlags()) are inherited by all subcommands; local flags (Flags()) apply only to the declaring command.

rootCmd.PersistentFlags().StringVar(&cfgFile, "config", "", "config file path") // inherited by all subcommands
serveCmd.Flags().IntVar(&port, "port", 8080, "listen port")                     // local to serveCmd only
serveCmd.MarkFlagRequired("port")
serveCmd.MarkFlagsMutuallyExclusive("json", "yaml")

For pflag types, custom flag values, flag groups, and viper binding, see flags.md.

Completions primer

Cobra generates shell completions automatically. Extend them with:

  • ValidArgs []string — static positional arg completion.
  • ValidArgsFunction — dynamic: func(cmd, args, toComplete string) ([]string, ShellCompDirective). Return ShellCompDirectiveNoFileComp to suppress file fallback.
  • RegisterFlagCompletionFunc(name, fn) — flag value completion.

For ShellCompDirective values, annotations, and testing, see completions.md.

Testing commands

Test commands by executing them programmatically. Never use os.Stdout / os.Stderr directly in command handlers — use cmd.OutOrStdout() / cmd.ErrOrStderr() so tests can redirect output.

func TestServeCmd(t *testing.T) {
    buf := new(bytes.Buffer)
    rootCmd.SetOut(buf)
    rootCmd.SetArgs([]string{"serve", "--port", "9090"})
    require.NoError(t, rootCmd.Execute())
    assert.Contains(t, buf.String(), "listening on :9090")
}

Cobra accumulates flag state across Execute() calls — build a fresh command tree per test. For isolation patterns, golden files, and testing completions, see testing.md.

Best Practices

  1. Always use RunE, never Run — Run cannot return an error; the only escape is os.Exit or panic, bypassing defers.
  2. Put config initialization in PersistentPreRunE — it runs before every subcommand; the right place for viper binding and auth checks.
  3. Validate positional args with Args, not inside RunE — Args gives cobra's standard error messages; MatchAll composes validators.
  4. Use cmd.OutOrStdout() / cmd.ErrOrStderr() for all output — direct os.Stdout writes cannot be captured by tests.
  5. Re-create the command tree per test — cobra accumulates flag state across Execute() calls on the same instance.

Common Mistakes

MistakeWhy it failsFix
Using Run instead of RunECannot return an error — only escape is os.Exit or panic, bypassing defersUse RunE — return the error, let cobra handle the exit
Writing len(args) checks in RunEBypasses cobra's standard error messages ("accepts 1 arg, received 2")Declare Args: cobra.ExactArgs(1) on the command
Writing to os.Stdout directlyTests cannot capture output — os-level file handles can't be redirectedUse cmd.OutOrStdout() / cmd.ErrOrStderr()
Child PersistentPreRunE silently drops parent'sCobra does not chain — the child replaces the parent's hook entirelyCall parent.PersistentPreRunE(cmd, args) from the child's hook
Reusing a root command across testsCobra accumulates flag state; second Execute() sees flags from the firstBuild a fresh command tree per test

Further Reading

  • commands-and-args.md — full PreRun*/PostRun* chain, every Args validator, PersistentPreRunE inheritance rules
  • flags.md — pflag types, required/exclusive/oneRequired groups, custom value types, viper binding
  • completions.md — ShellCompDirective set, annotation-based completions, testing completions
  • generators.md — man page, markdown, YAML, RST doc generation; cobra-cli scaffolder
  • testing.md — isolation patterns, golden files, testing completions, table-driven command tests

Cross-References

  • → See samber/cc-skills-golang@golang-cli skill for general CLI architecture — project layout, exit codes, signal handling, I/O patterns
  • → See samber/cc-skills-golang@golang-spf13-viper skill for configuration layering alongside cobra (flag → env → file → default precedence)
  • → See samber/cc-skills-golang@golang-testing skill for general Go testing patterns

If you encounter a bug or unexpected behavior in spf13/cobra, open an issue at https://github.com/spf13/cobra/issues.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.