name: add_platform.implement description: "Creates platform adapter, templates, tests with 100% coverage, and README documentation. Use after adding hook capabilities." user-invocable: false hooks: Stop: - hooks: - type: command command: ".deepwork/jobs/add_platform/hooks/run_tests.sh" SubagentStop: - hooks: - type: command command: ".deepwork/jobs/add_platform/hooks/run_tests.sh"

add_platform.implement

Step 3/4 in integrate workflow

Full workflow to integrate a new AI platform into DeepWork

Adds a new AI platform to DeepWork with adapter, templates, and tests. Use when integrating Cursor, Windsurf, or other AI coding tools.

Prerequisites (Verify First)

Before proceeding, confirm these steps are complete:

• /add_platform.research
• /add_platform.add_capabilities

Instructions

Goal: Creates platform adapter, templates, tests with 100% coverage, and README documentation. Use after adding hook capabilities.

Implement Platform Support

Objective

Create the complete platform implementation including the adapter class, command templates, comprehensive tests, and documentation updates.

Task

Build the full platform support by implementing the adapter, creating templates, writing tests with 100% coverage, and updating the README.

Prerequisites

Read the outputs from previous steps:

• doc/platforms/<platform_name>/cli_configuration.md - For template structure
• src/deepwork/schemas/job_schema.py - For current schema
• src/deepwork/adapters.py - For adapter patterns

Also review existing implementations for reference:

• src/deepwork/templates/claude/ - Example templates
• tests/ - Existing test patterns

Process

1. Create the platform adapter class

   Add a new adapter class to src/deepwork/adapters.py:

   class NewPlatformAdapter(PlatformAdapter):
       """Adapter for <Platform Name>."""
   
       platform_name = "<platform_name>"
       command_directory = "<path to commands>"  # e.g., ".cursor/commands"
       command_extension = ".md"  # or appropriate extension
   
       def get_hook_support(self) -> dict:
           """Return which hooks this platform supports."""
           return {
               "stop_hooks": True,  # or False/None
               # ... other hooks
           }
   
       def generate_command(self, step: StepDefinition, job: JobDefinition) -> str:
           """Generate command file content for this platform."""
           # Use Jinja2 template
           template = self.env.get_template(f"{self.platform_name}/command.md.j2")
           return template.render(step=step, job=job)
   

2. Create command templates

   Create templates in src/deepwork/templates/<platform_name>/:

   • command.md.j2 - Main command template
   • Any other templates needed for the platform's format

   Use the CLI configuration documentation to ensure the template matches the platform's expected format.

3. Register the adapter

   Update the adapter registry in src/deepwork/adapters.py:

   PLATFORM_ADAPTERS = {
       "claude": ClaudeAdapter,
       "<platform_name>": NewPlatformAdapter,
       # ... other adapters
   }
   

4. Write comprehensive tests

   Create tests in tests/ that cover:

   • Adapter instantiation
   • Hook support detection
   • Command generation
   • Template rendering
   • Edge cases (empty inputs, special characters, etc.)
   • Integration with the sync command

   Critical: Tests must achieve 100% coverage of new code.

5. Update README.md

   Add the new platform to README.md:

   • Add to "Supported Platforms" list
   • Add installation instructions:

     deepwork install --platform <platform_name>
     

   • Document any platform-specific notes or limitations

6. Run tests and verify coverage

   uv run pytest --cov=src/deepwork --cov-report=term-missing
   

   • All tests must pass
   • New code must have 100% coverage
   • If coverage is below 100%, add more tests

7. Iterate until tests pass with full coverage

   This step has a stop_hooks script that runs tests. Keep iterating until:

   • All tests pass
   • Coverage is 100% for new functionality

Output Format

templates/

Location: src/deepwork/templates/<platform_name>/

Create the following files:

command.md.j2:

{# Template for <platform_name> command files #}
{# Follows the platform's expected format from cli_configuration.md #}

[Platform-specific frontmatter or metadata]

# {{ step.name }}

{{ step.description }}

## Instructions

{{ step.instructions_content }}

[... rest of template based on platform format ...]

tests/

Location: tests/test_<platform_name>_adapter.py

"""Tests for the <platform_name> adapter."""
import pytest
from deepwork.adapters import NewPlatformAdapter

class TestNewPlatformAdapter:
    """Test suite for NewPlatformAdapter."""

    def test_adapter_initialization(self):
        """Test adapter can be instantiated."""
        adapter = NewPlatformAdapter()
        assert adapter.platform_name == "<platform_name>"

    def test_hook_support(self):
        """Test hook support detection."""
        adapter = NewPlatformAdapter()
        hooks = adapter.get_hook_support()
        assert "stop_hooks" in hooks
        # ... more assertions

    def test_command_generation(self):
        """Test command file generation."""
        # ... test implementation

    # ... more tests for 100% coverage

README.md

Add to the existing README.md:

## Supported Platforms

- **Claude Code** - Anthropic's CLI for Claude
- **<Platform Name>** - [Brief description]

## Installation

### <Platform Name>

```bash
deepwork install --platform <platform_name>

[Any platform-specific notes]

## Quality Criteria

- Platform adapter class added to `src/deepwork/adapters.py`:
  - Inherits from `PlatformAdapter`
  - Implements all required methods
  - Registered in `PLATFORM_ADAPTERS`
- Templates created in `src/deepwork/templates/<platform_name>/`:
  - `command.md.j2` exists and renders correctly
  - Format matches platform's expected command format
- Tests created in `tests/`:
  - Cover all new adapter functionality
  - Cover template rendering
  - All tests pass
- Test coverage is 100% for new code:
  - Run `uv run pytest --cov=src/deepwork --cov-report=term-missing`
  - No uncovered lines in new code
- README.md updated:
  - Platform listed in supported platforms
  - Installation command documented
  - Any platform-specific notes included
- When all criteria are met, include `<promise>✓ Quality Criteria Met</promise>` in your response

## Context

This is the core implementation step. The adapter you create will be responsible for:
- Determining where command files are placed
- Generating command file content from job definitions
- Handling platform-specific features and hooks

The templates use Jinja2 and should produce files that match exactly what the platform expects. Reference the CLI configuration documentation frequently to ensure compatibility.

## Tips

- Study the existing `ClaudeAdapter` as a reference implementation
- Run tests frequently as you implement
- Use `--cov-report=html` for a detailed coverage report
- If a test is hard to write, the code might need refactoring
- Template syntax errors often show up at runtime - test early


### Job Context

A workflow for adding support for a new AI platform (like Cursor, Windsurf, etc.) to DeepWork.

The **integrate** workflow guides you through four phases:
1. **Research**: Capture the platform's CLI configuration and hooks system documentation
2. **Add Capabilities**: Update the job schema and adapters with any new hook events
3. **Implement**: Create the platform adapter, templates, tests (100% coverage), and README updates
4. **Verify**: Ensure installation works correctly and produces expected files

The workflow ensures consistency across all supported platforms and maintains
comprehensive test coverage for new functionality.

**Important Notes**:
- Only hooks available on slash command definitions should be captured
- Each existing adapter must be updated when new hooks are added (typically with null values)
- Tests must achieve 100% coverage for any new functionality
- Installation verification confirms the platform integrates correctly with existing jobs


## Required Inputs


**Files from Previous Steps** - Read these first:
- `job_schema.py` (from `add_capabilities`)
- `adapters.py` (from `add_capabilities`)
- `cli_configuration.md` (from `research`)

## Work Branch

Use branch format: `deepwork/add_platform-[instance]-YYYYMMDD`

- If on a matching work branch: continue using it
- If on main/master: create new branch with `git checkout -b deepwork/add_platform-[instance]-$(date +%Y%m%d)`

## Outputs

**Required outputs**:
- `templates/` (directory)
- `tests/` (directory)
- `README.md`

## Guardrails

- Do NOT skip prerequisite verification if this step has dependencies
- Do NOT produce partial outputs; complete all required outputs before finishing
- Do NOT proceed without required inputs; ask the user if any are missing
- Do NOT modify files outside the scope of this step's defined outputs

**Validation script**: `.deepwork/jobs/add_platform/hooks/run_tests.sh` (runs automatically)
## On Completion

1. Verify outputs are created
2. Inform user: "integrate step 3/4 complete, outputs: templates/, tests/, README.md"
3. **Continue workflow**: Use Skill tool to invoke `/add_platform.verify`

---

**Reference files**: `.deepwork/jobs/add_platform/job.yml`, `.deepwork/jobs/add_platform/steps/implement.md`