---
description: "{{ one_line_description }}"
mode: subagent
---
# {{ name }}
{{ one_line_description }}
## When to use
Invoke this agent when the user's task matches any of these triggers:
{% for phrase in when_to_use_phrases %}- {{ phrase }}
{% endfor %}
## When NOT to use
- Do not use this agent if the user only wants to *read* or *inspect* {{ display_name }} without performing an action — prefer reading docs directly.
- Do not use this agent for tasks outside {{ category_hint }} it was built for.
## Installation
{% if has_cli %}
Ensure `{{ cli_binary }}` is installed and available on PATH before invoking the examples below.
{% else %}
This is a library, not a CLI. The agent uses it by importing from the host project, not by spawning a binary.
Import pattern:
```
{{ import_pattern }}
```
{% endif %}
{% if has_cli %}
## Invocation
The exact command an agent should run to use this tool:
```
{{ invocation_command }}
```
Run `{{ cli_binary }} --help` to discover the available flags. Documented flags (verified by `skillpack verify`):
{% for flag in documented_flags %}- `{{ flag }}`
{% endfor %}{% if documented_subcommands %}
### Subcommands
This CLI uses subcommands. Each one has its own flags (verified by `skillpack verify`):
{% for sub in documented_subcommands %}- `{{ sub.name }}`{%- if sub.flags %} — flags: {% for f in sub.flags %}`{{ f }}`{% if not loop.last %}, {% endif %}{% endfor %}{% endif %}
{% endfor %}
Run `{{ cli_binary }} <subcommand> --help` for a subcommand's full options.
{% endif %}{% else %}
## Usage
There is no CLI to invoke. Use the import pattern above and call the relevant API surface directly from the task.
{% endif %}
## Footguns / gotchas
- Verify the tool is installed before relying on it; do not assume availability.
{% if has_cli %}- Prefer the documented `{{ cli_binary }}` flags; undocumented flags may change without notice.
{% endif %}- This file is authoritative for agent invocation; if it drifts from the CLI's real `--help`, run `skillpack verify` to catch it.