marketplace/cli/internal/cmd/gen.go

package cmd

import (
	"errors"
	"flag"
	"fmt"
	"io/fs"
	"os"
	"path/filepath"
	"strings"
	"unicode/utf8"
)

// The region markers. What sits between them belongs to the generator; the
// rest of the file belongs to whoever wrote it.
const (
	genOpen  = "<!-- kettle:gen -->"
	genClose = "<!-- /kettle:gen -->"
)

// genBanner opens every generated region. The first thing anybody who finds
// the block wants to do is edit it in place, so the block says who wrote it and
// which command writes it again.
const genBanner = "**Generated from the kettle command registry by `kettle gen skills`.** " +
	"Everything between the two markers is replaced on the next run — " +
	"hand-written prose belongs outside them."

// exampleAlign is the widest example command that still gets its `# what`
// padded into a column. One long pipeline would otherwise push every other
// comment off the right edge of the page.
const exampleAlign = 56

func init() {
	register(&Command{
		Name:  "gen",
		Group: GroupProject,
		Args:  "skills",
		Short: "write the plugin's SKILL.md files from the command registry",
		Long: `A SKILL.md tells an agent how to invoke this binary. Hand-written, it drifts: a
flag is renamed here and the documentation goes on recommending the old one,
and the agent that reads it fails in a way nobody traces back to a stale
sentence. Everything those files say about a command — its usage line, its
flags with their defaults, its worked examples — is already in the registry
this binary is built from, so it is written from there and cannot disagree.

THE GENERATOR OWNS A REGION, NOT A FILE. Each SKILL.md carries a pair of HTML
comment markers — ` + "`kettle:gen`" + ` to open and ` + "`/kettle:gen`" + ` to close, both written in
the ` + "`<!-- … -->`" + ` form and visible at the top and bottom of the block below.
Everything between them is replaced on every run; every byte outside them comes
back exactly as it was, which matters most for ` + "`description:`" + `, the prose that
decides whether an agent loads the skill at all, and the one thing here that no
generator can write.

A file with no markers is REPORTED AND LEFT ALONE, never overwritten: clobbering
somebody's prose because they forgot a marker is the failure this design exists
to prevent. A file that does not exist yet is created with a frontmatter stub
around a generated block, for a human to fill in.

The output is deterministic to the byte — no timestamps, no map iteration — so
regenerating something that has not changed produces no diff. --check is that
property made useful: it writes nothing and exits 1 when any file on disk
differs from what would be generated, which is what a pre-commit hook or a CI
step calls. It wins over --dry-run when both are given.`,
		Examples: []Example{
			{"kettle gen skills --out ../plugins/tea/skills", "write the region in every group's SKILL.md"},
			{"kettle gen skills --out ../plugins/tea/skills --dry-run", "print what would change; write nothing"},
			{"kettle gen skills --out ../plugins/tea/skills --check", "exit 1 if the docs are out of date"},
		},
		Setup: func(fs *flag.FlagSet) func([]string) error {
			out := fs.String("out", "", "directory the skills live in; one <group>/SKILL.md under it")
			dryRun := fs.Bool("dry-run", false, "print what would change; write nothing")
			check := fs.Bool("check", false, "write nothing, exit 1 if anything is out of date")

			return func(args []string) error {
				target := "skills"
				if len(args) > 0 {
					target = args[0]
				}
				if len(args) > 1 || target != "skills" {
					return Fail("the only target is `skills` — try `kettle gen skills --out <dir>`")
				}
				if *out == "" {
					return Fail("--out is required — the directory the SKILL.md files live under")
				}
				return genSkills(*out, *dryRun, *check)
			}
		},
	})
}

// errNoRegion is what a file that the generator may not touch reports.
var errNoRegion = errors.New("no " + genOpen + " … " + genClose + " region")

func genSkills(dir string, dryRun, check bool) error {
	// --check is a read-only question about the working tree, so it overrules
	// --dry-run rather than combining with it.
	if check {
		dryRun = true
	}

	groups := docGroups()
	var written, unchanged, outdated, kept int
	for _, group := range groups {
		path := filepath.Join(dir, group, "SKILL.md")
		block, err := renderGroup(commandsIn(group))
		if err != nil {
			return err
		}

		existing, err := os.ReadFile(path)
		switch {
		case errors.Is(err, fs.ErrNotExist):
			outdated++
			if check {
				fmt.Printf("%-13s %s\n", "missing", path)
				continue
			}
			if dryRun {
				fmt.Printf("%-13s %s\n", "would create", path)
				continue
			}
			if err := writeFile(path, stubFile(group, block)); err != nil {
				return err
			}
			written++
			fmt.Printf("%-13s %s\n", "created", path)

		case err != nil:
			return err

		default:
			want, err := spliceRegion(string(existing), block)
			if err != nil {
				// Reported, never repaired: a missing marker is somebody's
				// prose sitting where the block used to be.
				kept++
				fmt.Fprintf(os.Stderr, "kettle gen: %s left alone — %v\n", path, err)
				continue
			}
			if want == string(existing) {
				unchanged++
				fmt.Printf("%-13s %s\n", "unchanged", path)
				continue
			}
			outdated++
			if check {
				fmt.Printf("%-13s %s\n", "stale", path)
				continue
			}
			if dryRun {
				fmt.Printf("%-13s %s\n", "would update", path)
				continue
			}
			if err := writeFile(path, want); err != nil {
				return err
			}
			written++
			fmt.Printf("%-13s %s\n", "updated", path)
		}
	}

	switch {
	case check:
		fmt.Printf("%d file(s) checked, %d out of date, %d without a region\n",
			len(groups), outdated, kept)
		if outdated > 0 {
			fmt.Printf("run `kettle gen skills --out %s`\n", dir)
			return SilentError{Code: 1}
		}
	case dryRun:
		fmt.Printf("%d file(s) would change, %d unchanged, %d without a region — nothing was written\n",
			outdated, unchanged, kept)
	default:
		fmt.Printf("%d file(s) written, %d unchanged, %d without a region\n", written, unchanged, kept)
	}
	return nil
}

// docGroups lists the groups that have commands, in the order Commands()
// returns them — the same order twice, so two runs cannot differ.
func docGroups() []string {
	var out []string
	seen := map[string]bool{}
	for _, c := range Commands() {
		if c.Group == "" {
			fmt.Fprintf(os.Stderr, "kettle gen: command %q has no group and is in no skill\n", c.Name)
			continue
		}
		if !seen[c.Group] {
			seen[c.Group] = true
			out = append(out, c.Group)
		}
	}
	return out
}

func commandsIn(group string) []*Command {
	var out []*Command
	for _, c := range Commands() {
		if c.Group == group {
			out = append(out, c)
		}
	}
	return out
}

// renderGroup is the generated block for one group, without the markers and
// without a trailing newline.
func renderGroup(cmds []*Command) (string, error) {
	var b strings.Builder
	b.WriteString(genBanner)
	b.WriteString("\n")

	for _, c := range cmds {
		text := renderCommand(c)
		// A block holding either marker would cut itself in half on the next
		// run — the splice would end the region in the middle of the prose that
		// mentions it. Loud here rather than quietly truncated on disk.
		if strings.Contains(text, genOpen) || strings.Contains(text, genClose) {
			return "", Fail("command %q spells a region marker out in full; the generated block would then end inside itself — write it another way", c.Name)
		}
		b.WriteString(text)
	}
	return strings.TrimRight(b.String(), "\n"), nil
}

func renderCommand(c *Command) string {
	var b strings.Builder
	fmt.Fprintf(&b, "\n## `%s`\n\n%s\n", c.Usage(), c.Short)
	if long := strings.TrimSpace(c.Long); long != "" {
		b.WriteString("\n" + long + "\n")
	}
	if flags := c.Flags(); len(flags) > 0 {
		b.WriteString("\n| flag | default | what it does |\n| --- | --- | --- |\n")
		for _, f := range flags {
			fmt.Fprintf(&b, "| `--%s` | %s | %s |\n", f.Name, defaultCell(f.DefValue), cell(f.Usage))
		}
	}
	if len(c.Examples) > 0 {
		w := exampleWidth(c.Examples)
		b.WriteString("\n```bash\n")
		for _, e := range c.Examples {
			pad := w - utf8.RuneCountInString(e.Cmd)
			if pad < 0 {
				pad = 0
			}
			fmt.Fprintf(&b, "%s%s  # %s\n", e.Cmd, strings.Repeat(" ", pad), e.What)
		}
		b.WriteString("```\n")
	}
	return b.String()
}

// exampleWidth is the column the `# what` comments line up at. Runes, not
// bytes: an example with Cyrillic in it would otherwise pull the column left by
// however many multi-byte characters it holds.
func exampleWidth(examples []Example) int {
	w := 0
	for _, e := range examples {
		if n := utf8.RuneCountInString(e.Cmd); n > w && n <= exampleAlign {
			w = n
		}
	}
	return w
}

func defaultCell(v string) string {
	if v == "" {
		return "—"
	}
	return "`" + cell(v) + "`"
}

// cell keeps a value from breaking out of its table row.
func cell(s string) string {
	s = strings.ReplaceAll(s, "\n", " ")
	return strings.ReplaceAll(s, "|", `\|`)
}

func region(block string) string {
	return genOpen + "\n" + block + "\n" + genClose
}

// spliceRegion swaps the block into existing, leaving every other byte alone.
func spliceRegion(existing, block string) (string, error) {
	start := strings.Index(existing, genOpen)
	if start < 0 {
		return "", errNoRegion
	}
	rest := start + len(genOpen)
	end := strings.Index(existing[rest:], genClose)
	if end < 0 {
		return "", fmt.Errorf("%s is missing its %s", genOpen, genClose)
	}
	return existing[:start] + region(block) + existing[rest+end+len(genClose):], nil
}

// stubFile is a new SKILL.md: the least frontmatter that is still a skill,
// and the region.
//
// The description is left as a TODO on purpose. It is the sentence that decides
// whether an agent loads this skill at all — prose a human tunes against real
// failures to trigger, and the one thing here a generator has no way to write.
func stubFile(group, block string) string {
	title := "# kettle " + group + "\n"
	if blurb := groupBlurb[group]; blurb != "" {
		title += "\n" + blurb + "\n"
	}
	return "---\n" +
		"name: " + group + "\n" +
		"description: TODO — write this by hand. It is the only thing that decides whether an agent loads this skill at all, so it is prose a human tunes; kettle gen never reads or writes it.\n" +
		"---\n\n" +
		title + "\n" +
		region(block) + "\n"
}

func writeFile(path, content string) error {
	if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
		return err
	}
	return os.WriteFile(path, []byte(content), 0o644)
}