Generate a BSage Skill Boilerplate

VerifiedSafe

Creates a new BSage Skill with standard structure and boilerplate files.

Sby Skills Guide Bot
DevelopmentIntermediate
007/24/2026
Claude Code
#scaffolding#boilerplate#code-generation#skill-development

Recommended for

Our review

Automatically generates the standard boilerplate for a new BSage skill, including YAML, Python, and test files.

Strengths

  • Reduces startup time for creating a new skill
  • Provides category-appropriate default configurations (input, process, output, meta)
  • Includes a unit test skeleton automatically

Limitations

  • Generated code is generic and requires custom implementation
  • Does not handle external dependency validation
When to use it

When you need to quickly scaffold a new BSage skill with a consistent structure.

When not to use it

When modifying an existing skill or when your project does not use the BSage framework.

Security analysis

Safe
Quality score88/100

This skill is a meta-template generator that creates boilerplate files; it does not execute any system commands, network operations, or destructive actions. The provided examples are static and instructional, with no risk of code injection or exfiltration.

No concerns found

Examples

Create input skill
/skill input/calendar-input
Create process skill
/skill process/garden-writer
Create output skill
/skill output/git-output

name: skill description: Create a new BSage Skill with standard boilerplate

Skill Command

Generate a new Skill with standard structure and boilerplate.

Usage

/skill <category>/<skill-name>

Example

/skill input/calendar-input
/skill process/garden-writer
/skill output/git-output
/skill meta/skill-builder

What This Does

  1. Creates skill directory at skills/<skill-name>/
  2. Generates skill.yaml with category-appropriate defaults
  3. Generates skill.py with execute(context) boilerplate
  4. Creates test file at bsage/tests/test_skill_<name>.py

Generated Structure

For /skill input/calendar-input:

skills/
└── calendar-input/
    ├── skill.yaml      # Metadata + trigger config
    └── skill.py         # execute(context) function

bsage/
└── tests/
    └── test_skill_calendar_input.py

Generated Files

skills/calendar-input/skill.yaml

name: calendar-input
version: 1.0.0
author: bslab
category: input
is_dangerous: false
entrypoint: skill.py::execute
trigger:
  type: cron
  schedule: "*/15 * * * *"
rules:
  - garden-writer
description: Google Calendar 일정을 2nd Brain으로 수집

skills/calendar-input/skill.py

from __future__ import annotations


async def execute(context):
    """Google Calendar 일정을 수집하여 seed로 저장."""
    context.logger.info("calendar_input_start")

    creds = await context.credentials.get("google-calendar")
    # Skill 내부에서 직접 API 연결 처리
    events = await fetch_calendar_events(creds)

    await context.garden.write_seed("calendar", {"events": events})

    context.logger.info("calendar_input_complete", event_count=len(events))
    return {"collected": len(events)}

bsage/tests/test_skill_calendar_input.py

import pytest
from unittest.mock import AsyncMock, MagicMock


@pytest.fixture
def mock_context():
    context = MagicMock()
    context.logger = MagicMock()
    context.credentials = MagicMock()
    context.credentials.get = AsyncMock(return_value={"client_id": "test"})
    context.garden = AsyncMock()
    context.garden.write_seed = AsyncMock()
    return context


@pytest.mark.asyncio
async def test_execute_collects_events(mock_context):
    from importlib.util import spec_from_file_location, module_from_spec
    spec = spec_from_file_location("skill", "skills/calendar-input/skill.py")
    module = module_from_spec(spec)
    spec.loader.exec_module(module)

    result = await module.execute(mock_context)

    assert result["collected"] >= 0
    mock_context.garden.write_seed.assert_called_once()

Category Templates

InputSkill

category: input
trigger:
  type: cron
  schedule: "*/15 * * * *"
rules:
  - garden-writer

ProcessSkill

category: process
is_dangerous: false    # true if external side effects

OutputSkill

category: output

MetaSkill

category: meta
is_dangerous: false

Next Steps

After generation:

  1. Fill in skill.py implementation
  2. Update skill.yaml fields (rules, trigger)
  3. Add test cases for error paths
  4. Run tests with /test
  5. Verify Skill loads correctly with SkillLoader
Related skills