get_plugin_dependency_graph

Instructions

Implementation Reference

• src/mcp_jenkins/server/plugin.py:78-90 (handler)

  The MCP tool handler function that exposes get_plugin_dependency_graph as a 'read' tool. Delegates to the REST client.

  async def get_plugin_dependency_graph(ctx: Context, short_name: str) -> dict:
      """Get dependency graph for a specific plugin in Graphviz format
  
      Recursively analyzes dependencies down to leaf nodes.
      Returns nodes and edges that can be used to generate a dependency graph.
  
      Args:
          short_name: The short name of the plugin to analyze
  
      Returns:
          A dictionary with 'nodes' and 'edges' for Graphviz rendering
      """
      return jenkins(ctx).get_plugin_dependency_graph(short_name=short_name)

• src/mcp_jenkins/jenkins/rest_client.py:851-898 (handler)

  The core implementation of get_plugin_dependency_graph. Recursively traverses plugin dependencies, building nodes and edges for a Graphviz graph.

  def get_plugin_dependency_graph(self, short_name: str) -> dict:
      """Get dependency graph for a specific plugin in Graphviz format.
  
      Recursively analyzes dependencies down to leaf nodes (plugins with no dependencies).
  
      Args:
          short_name: The short name of the plugin to analyze.
  
      Returns:
          A dictionary containing 'nodes' and 'edges' for Graphviz rendering.
      """
      plugins = self.get_plugins(depth=2)
      installed = {p['shortName']: p for p in plugins}
  
      if short_name not in installed:
          return {'nodes': [], 'edges': [], 'error': f'Plugin not found: {short_name}'}
  
      nodes = []
      edges = []
      visited = set()
  
      def traverse(name: str) -> None:
          if name in visited:
              return
          visited.add(name)
  
          if name not in installed:
              nodes.append({'id': name, 'label': name, 'status': 'missing'})
              return
  
          plugin = installed[name]
          nodes.append(
              {
                  'id': name,
                  'label': f'{name}\n({plugin.get("version", "?")})',
                  'status': 'installed',
              }
          )
  
          deps = plugin.get('dependencies', [])
          for dep in deps:
              dep_name = dep.get('shortName', '')
              edges.append({'from': name, 'to': dep_name})
              traverse(dep_name)
  
      traverse(short_name)
  
      return {'nodes': nodes, 'edges': edges}

• src/mcp_jenkins/server/__init__.py:30-34 (registration)

  The MCP server instance (JenkinsMCP) is created here. The @mcp.tool decorator in plugin.py registers the tool via the 'mcp' instance imported from this module.

  mcp = JenkinsMCP('mcp-jenkins', lifespan=lifespan)
  
  # Import tool modules to register them with the MCP server
  # This must happen after mcp is created so the @mcp.tool() decorators can reference it
  from mcp_jenkins.server import build, item, node, plugin, queue, view  # noqa: F401, E402

• tests/test_server/test_plugin.py:66-82 (helper)

  Test that validates get_plugin_dependency_graph returns correct nodes and edges for a plugin with dependencies.

  async def test_get_plugin_dependency_graph(mock_jenkins, mocker):
      mock_jenkins.get_plugin_dependency_graph.return_value = {
          'nodes': [
              {'id': 'plugin-a', 'label': 'plugin-a\n(1.0)', 'status': 'installed'},
              {'id': 'dep-a', 'label': 'dep-a\n(1.0)', 'status': 'installed'},
          ],
          'edges': [{'from': 'plugin-a', 'to': 'dep-a'}],
      }
  
      result = await plugin.get_plugin_dependency_graph(mocker.Mock(), short_name='plugin-a')
      assert result == {
          'nodes': [
              {'id': 'plugin-a', 'label': 'plugin-a\n(1.0)', 'status': 'installed'},
              {'id': 'dep-a', 'label': 'dep-a\n(1.0)', 'status': 'installed'},
          ],
          'edges': [{'from': 'plugin-a', 'to': 'dep-a'}],
      }