10. Skill arguments

⋯ ---name: feeddescription: Feeds an animal something, demonstrating skill arguments.--- Respond with exactly these two lines and nothing else: Feed the $0 some $1.Full request: $ARGUMENTS Do not call any tools.

⋯ ---name: rabbitdescription: Reports the burrow location for the project's rabbit mascot.--- Respond with exactly one word: burrow Do not add any other text, punctuation, or formatting.

⋯ """Stage 9: slash commands, level 2 (invoke) A prompt starting with "/" invokes a skill by name: the first word afterthe slash resolves to .petite/skills/<name>/SKILL.md, and that skill'sbody (everything after the closing "---") replaces the raw prompt as theuser message. Only the invoked skill's body is ever loaded — every otherskill stays at level 1 (name + description in the system prompt), so themodel never sees two conflicting sets of instructions at once.Stage 10: skill arguments ($ARGUMENTS, $0, $1, ...) A slash command can pass arguments after the skill name:"/feed rabbit carrots" invokes "feed" with args ["rabbit", "carrots"].The body can reference them with placeholders, substituted before thebody is sent to the model:   $ARGUMENTS         everything after the skill name, joined by spaces  $ARGUMENTS[0], [1] a single argument, by position  $0, $1, ...        shorthand for the same thing Order matters when substituting: $ARGUMENTS[0] must be handled before thebare $ARGUMENTS, or a naive string replace would mangle it into"rabbit carrots[0]".""" import osimport re import yamlfrom pydantic import BaseModel, Field, field_validator⋯ def load_skill_body(name, skills_dir=SKILLS_DIR):  def resolve_slash_command(prompt, skills_dir=SKILLS_DIR):    """If prompt starts with '/', resolve the first word as a skill name    and return that skill's body to use as the user message content.    """If prompt starts with '/', resolve the first word as a skill name,    substitute any argument placeholders in its body, and return that as    the user message content.     Returns None if the prompt isn't a slash command or the named skill    can't be found, so the caller falls back to the raw prompt.⋯ def resolve_slash_command(prompt, skills_dir=SKILLS_DIR):    if not rest:        return None     skill_name = rest.split()[0]    return load_skill_body(skill_name, skills_dir)    parts = rest.split()    skill_name, args = parts[0], parts[1:]     body = load_skill_body(skill_name, skills_dir)    if body is None:        return None     return substitute_arguments(body, args)  def substitute_arguments(body, args):    """Replace $ARGUMENTS, $ARGUMENTS[n] and $n placeholders in a skill    body with the given positional arguments.     Missing indices substitute to an empty string rather than raising.    """     def arg_at(index):        return args[index] if 0 <= index < len(args) else ""     # $ARGUMENTS[n] first — a bare $ARGUMENTS replace would otherwise    # mangle "$ARGUMENTS[0]" into "<joined args>[0]".    body = re.sub(r"\$ARGUMENTS\[(\d+)\]", lambda m: arg_at(int(m.group(1))), body)    body = body.replace("$ARGUMENTS", " ".join(args))    body = re.sub(r"\$(\d+)", lambda m: arg_at(int(m.group(1))), body)     return body