skill-name

name	{"[object Object]":null}
description	{"[object Object]":null}

{{TITLE}}

{{INTRO_PARAGRAPH}}

Input

Arguments: $ARGUMENTS

The user may provide optional arguments when invoking this skill. Dynamic context is available via:

• $ARGUMENTS — raw argument string passed by the user
• !cmd — command output injected before Claude reads the prompt (e.g., `!`git status -s)
  • Injections run at parse time, before $ARGUMENTS is available — never inject a command containing a placeholder you expect to be substituted.
  • A non-zero exit aborts skill load; it does not degrade to empty output. Guard anything that can fail: !git status -s 2>/dev/null || echo "(not a git repository)"``

Instructions

Phase 1: {{PHASE_1_TITLE}}

1. {{STEP_1}}
2. {{STEP_2}}
3. {{STEP_3}}

Phase 2: {{PHASE_2_TITLE}}

1. {{STEP_1}}
2. {{STEP_2}}

Output

Describe what output the skill produces (in-conversation, file, etc.)

Example

User: /{{SKILL_NAME}}

Claude: [Example response]

Error Handling

• If {{ERROR_CONDITION_1}}: {{ERROR_ACTION_1}}
• If {{ERROR_CONDITION_2}}: {{ERROR_ACTION_2}}