orca.ai/pkg/skill/parser.go

package skill

import (
	"fmt"
	"os"
	"path/filepath"
	"sort"
	"strings"
)

// FrontmatterDelimiters for YAML frontmatter in SKILL.md files.
const (
	frontmatterDelim = "---"
)

// ParseSkillFile parses a SKILL.md file and returns a populated Skill struct.
//
// The expected format is:
//
//	---
//	name: my-skill
//	description: Does something useful
//	triggers: ["keyword1", "keyword2"]
//	---
//
//	# My Skill
//
//	Detailed description...
func ParseSkillFile(path string) (*Skill, error) {
	data, err := os.ReadFile(path)
	if err != nil {
		return nil, fmt.Errorf("skill: cannot read %q: %w", path, err)
	}

	return ParseSkillData(path, data)
}

// ParseSkillData parses SKILL.md content from raw bytes.
// The path parameter is used to locate the scripts/ directory.
func ParseSkillData(path string, data []byte) (*Skill, error) {
	content := string(data)

	skill := &Skill{
		Path:  path,
		Body:  content,
		Triggers: []string{},
	}

	// Parse YAML frontmatter
	rest, err := parseFrontmatter(content, skill)
	if err != nil {
		return nil, err
	}
	skill.Body = strings.TrimSpace(rest)

	// Validate required fields
	if skill.Name == "" {
		return nil, fmt.Errorf("skill: %q is missing 'name' in frontmatter", path)
	}

	// Discover scripts directory
	skillDir := filepath.Dir(path)
	scriptsDir := filepath.Join(skillDir, "scripts")
	skill.ScriptsDir = scriptsDir

	if info, err := os.Stat(scriptsDir); err == nil && info.IsDir() {
		scripts, err := discoverScripts(scriptsDir)
		if err != nil {
			return nil, fmt.Errorf("skill: failed to discover scripts in %q: %w", scriptsDir, err)
		}
		skill.Scripts = scripts
	}

	return skill, nil
}

// parseFrontmatter extracts YAML frontmatter delimited by "---" lines
// and populates the Skill struct fields.
func parseFrontmatter(content string, skill *Skill) (string, error) {
	content = strings.TrimSpace(content)

	if !strings.HasPrefix(content, frontmatterDelim) {
		// No frontmatter — treat entire content as body
		return content, nil
	}

	// Find the closing delimiter
	rest := content[len(frontmatterDelim):]
	rest = strings.TrimLeft(rest, "\n\r")

	endIdx := strings.Index(rest, "\n"+frontmatterDelim)
	if endIdx < 0 {
		// Also check for end-of-file style
		endIdx = strings.Index(rest, frontmatterDelim)
		if endIdx < 0 {
			return "", fmt.Errorf("skill: unclosed frontmatter in skill file")
		}
	}

	frontmatter := rest[:endIdx]
	body := rest[endIdx+len(frontmatterDelim)+1:]

	// Parse the YAML frontmatter (simple key-value parser)
	if err := parseSimpleYAML(frontmatter, skill); err != nil {
		return "", err
	}

	return body, nil
}

// parseSimpleYAML parses a simplified YAML format for skill frontmatter.
// Supports: string values, quoted strings, and array values.
func parseSimpleYAML(yaml string, skill *Skill) error {
	lines := strings.Split(yaml, "\n")
	for _, line := range lines {
		line = strings.TrimSpace(line)
		if line == "" || strings.HasPrefix(line, "#") {
			continue
		}

		colonIdx := strings.Index(line, ":")
		if colonIdx < 0 {
			continue
		}

		key := strings.TrimSpace(line[:colonIdx])
		value := strings.TrimSpace(line[colonIdx+1:])

		switch key {
		case "name":
			skill.Name = trimQuotes(value)

		case "description":
			skill.Description = trimQuotes(value)

		case "triggers":
			triggers, err := parseYAMLArray(value)
			if err != nil {
				return fmt.Errorf("skill: invalid triggers format: %w", err)
			}
			skill.Triggers = triggers
		}
	}
	return nil
}

// parseYAMLArray parses a YAML array like '["a", "b", "c"]' or '[a, b, c]'.
func parseYAMLArray(value string) ([]string, error) {
	value = strings.TrimSpace(value)

	// Handle YAML list format: ["a", "b"] or [a, b]
	if strings.HasPrefix(value, "[") && strings.HasSuffix(value, "]") {
		inner := value[1 : len(value)-1]
		if strings.TrimSpace(inner) == "" {
			return []string{}, nil
		}
		parts := splitCommas(inner)
		result := make([]string, len(parts))
		for i, p := range parts {
			result[i] = trimQuotes(strings.TrimSpace(p))
		}
		return result, nil
	}

	// Handle YAML list format with dashes:
	// triggers:
	//   - browser
	//   - navigate
	// (This would be handled line-by-line in a different flow)
	// For now, treat single value as a one-element list
	if value != "" && value != "[]" {
		return []string{trimQuotes(value)}, nil
	}

	return []string{}, nil
}

// splitCommas splits a comma-separated string respecting quoted sections.
func splitCommas(s string) []string {
	var parts []string
	var current strings.Builder
	inQuote := false
	quoteChar := byte(0)

	for i := 0; i < len(s); i++ {
		c := s[i]
		if inQuote {
			current.WriteByte(c)
			if c == quoteChar {
				inQuote = false
			}
		} else if c == '"' || c == '\'' {
			current.WriteByte(c)
			inQuote = true
			quoteChar = c
		} else if c == ',' {
			parts = append(parts, current.String())
			current.Reset()
		} else {
			current.WriteByte(c)
		}
	}
	if current.Len() > 0 {
		parts = append(parts, current.String())
	}
	return parts
}

// trimQuotes removes surrounding quotes from a string value.
func trimQuotes(s string) string {
	s = strings.TrimSpace(s)
	if len(s) >= 2 {
		if (s[0] == '"' && s[len(s)-1] == '"') || (s[0] == '\'' && s[len(s)-1] == '\'') {
			return s[1 : len(s)-1]
		}
	}
	return s
}

// discoverScripts lists all executable/readable files in a scripts directory.
func discoverScripts(scriptsDir string) ([]string, error) {
	entries, err := os.ReadDir(scriptsDir)
	if err != nil {
		return nil, err
	}

	var scripts []string
	for _, entry := range entries {
		if entry.IsDir() {
			continue
		}
		scripts = append(scripts, entry.Name())
	}

	sort.Strings(scripts)
	return scripts, nil
}

// LoadSkillFromDir loads a skill from a directory containing a SKILL.md file.
func LoadSkillFromDir(dir string) (*Skill, error) {
	skillPath := filepath.Join(dir, "SKILL.md")
	if _, err := os.Stat(skillPath); os.IsNotExist(err) {
		return nil, fmt.Errorf("skill: no SKILL.md found in %q", dir)
	}
	return ParseSkillFile(skillPath)
}