@sweny-ai/mcp
v0.1.2
Published
MCP server exposing SWEny workflows to Claude Code and Claude Desktop
Maintainers
Readme
@sweny-ai/mcp
MCP server that exposes SWEny workflows to Claude Code and Claude Desktop.
What it does
This package runs as a Model Context Protocol server over stdio, giving Claude two tools:
| Tool | Description |
|------|-------------|
| sweny_list_workflows | List built-in and custom workflows in the current project |
| sweny_run_workflow | Execute a triage or implement workflow and return structured results |
This lets Claude Code delegate complex multi-step tasks to SWEny's DAG executor, which manages context boundaries across nodes — something a single conversation context can't do well.
Setup
Claude Code
Add to your project's .claude/settings.json:
{
"mcpServers": {
"sweny": {
"command": "npx",
"args": ["-y", "@sweny-ai/mcp"]
}
}
}Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"sweny": {
"command": "npx",
"args": ["-y", "@sweny-ai/mcp"]
}
}
}Prerequisites
- A
.sweny.ymlconfig file in the working directory (runsweny initto create one) - Credentials for your configured providers (set via environment variables or
.env) @sweny-ai/coreinstalled (provides theswenyCLI binary)
Tools
sweny_list_workflows
Lists available workflows. Returns built-in workflows (triage, implement, seed-content) and any custom workflows found in .sweny/workflows/*.yml.
Parameters:
| Name | Type | Required | Description |
|------|------|----------|-------------|
| cwd | string | no | Working directory to search. Defaults to process.cwd(). |
sweny_run_workflow
Executes a SWEny workflow by spawning the sweny CLI. Returns structured JSON results on completion.
Parameters:
| Name | Type | Required | Description |
|------|------|----------|-------------|
| workflow | "triage" | "implement" | yes | Which workflow to run |
| input | string | for implement | Issue ID or URL (required for implement, ignored for triage) |
| cwd | string | no | Working directory containing .sweny.yml |
| dryRun | boolean | no | Skip side effects (no issues/PRs created) |
Notes:
- Triage discovers alerts automatically via your configured observability provider
- Implement requires an issue ID from your configured issue tracker
- Workflows can take several minutes to complete
- A 10-minute timeout kills hanging processes
Development
npm run build # Compile TypeScript + chmod +x
npm run typecheck # Type-check without emitting
npm test # Run all tests