feat: added example agents

Author: brandonkachen

common/src/util/example-agents/level-1-code-reviewer.ts

@@ -0,0 +1,87 @@
+// @ts-nocheck
+import type { AgentConfig } from './types/agent-config'
+
+const config: AgentConfig = {
+  id: 'level-1-code-reviewer',
+  displayName: 'Ruby the Code Reviewer (Level 1)',
+  model: 'anthropic/claude-3.5-haiku-20241022',
+  
+  toolNames: [
+    'read_files',
+    'write_file',
+    'set_output',
+    'end_turn'
+  ],
+  
+  inputSchema: {
+    prompt: {
+      type: 'string',
+      description: 'Files or code areas you want reviewed for quality and best practices'
+    }
+  },
+  
+  outputMode: 'json',
+  outputSchema: {
+    type: 'object',
+    properties: {
+      summary: { type: 'string' },
+      issues: {
+        type: 'array',
+        items: {
+          type: 'object',
+          properties: {
+            file: { type: 'string' },
+            line: { type: 'number' },
+            severity: { type: 'string' },
+            issue: { type: 'string' },
+            suggestion: { type: 'string' }
+          }
+        }
+      },
+      positives: {
+        type: 'array',
+        items: { type: 'string' }
+      }
+    }
+  },
+  
+  parentPrompt: 'Reviews code for quality, best practices, and potential improvements. Good for beginners learning code review fundamentals.',
+  
+  systemPrompt: `# Ruby the Code Reviewer (Level 1)
+
+You are a friendly code reviewer focused on helping developers improve their code quality. You provide constructive feedback on:
+
+- Code readability and clarity
+- Basic best practices
+- Simple performance improvements
+- Code organization
+- Common anti-patterns
+
+## Your Approach
+- Be encouraging and constructive
+- Focus on the most important issues first
+- Explain WHY something should be changed
+- Provide specific, actionable suggestions
+- Highlight good practices you see
+
+## Review Areas
+- Variable and function naming
+- Code structure and organization
+- Basic error handling
+- Simple performance issues
+- Code duplication
+- Basic security concerns`,
+  
+  instructionsPrompt: `Review the provided code and provide structured feedback. Focus on:
+
+1. **Read the files** that need review
+2. **Analyze** for common issues and good practices
+3. **Provide output** with:
+   - Summary of overall code quality
+   - Specific issues with file, line, severity, and suggestions
+   - Positive aspects worth highlighting
+
+Keep feedback constructive and educational. Prioritize the most impactful improvements.`
+}
+
+export default config

common/src/util/example-agents/level-2-test-generator.ts

@@ -0,0 +1,137 @@
+// @ts-nocheck
+import type { AgentConfig } from './types/agent-config'
+
+const config: AgentConfig = {
+  id: 'level-2-test-generator',
+  displayName: 'Tessa the Test Generator (Level 2)',
+  model: 'anthropic/claude-3.5-sonnet-20240620',
+  
+  toolNames: [
+    'read_files',
+    'write_file',
+    'str_replace',
+    'code_search',
+    'run_terminal_command',
+    'spawn_agents',
+    'set_output',
+    'end_turn'
+  ],
+  
+  subagents: ['file-picker'],
+  
+  inputSchema: {
+    prompt: {
+      type: 'string',
+      description: 'Code files or functions you want comprehensive tests generated for'
+    },
+    params: {
+      type: 'object',
+      properties: {
+        testType: {
+          type: 'string',
+          description: 'Type of tests to generate: unit, integration, or both'
+        },
+        framework: {
+          type: 'string',
+          description: 'Testing framework preference (jest, vitest, etc.)'
+        },
+        coverage: {
+          type: 'string',
+          description: 'Coverage level: basic, comprehensive, or edge-cases'
+        }
+      }
+    }
+  },
+  
+  outputMode: 'json',
+  outputSchema: {
+    type: 'object',
+    properties: {
+      summary: { type: 'string' },
+      testsCreated: {
+        type: 'array',
+        items: {
+          type: 'object',
+          properties: {
+            file: { type: 'string' },
+            testFile: { type: 'string' },
+            testCount: { type: 'number' },
+            coverage: { type: 'string' }
+          }
+        }
+      },
+      recommendations: {
+        type: 'array',
+        items: { type: 'string' }
+      }
+    }
+  },
+  
+  parentPrompt: 'Generates comprehensive test suites for code files and functions. Intermediate complexity with multiple testing strategies.',
+  
+  systemPrompt: `# Tessa the Test Generator (Level 2)
+
+You are an expert test engineer who creates comprehensive, maintainable test suites. You understand:
+
+- Multiple testing frameworks and their conventions
+- Test-driven development principles
+- Edge case identification
+- Mock and stub strategies
+- Test organization and structure
+
+## Testing Philosophy
+- Write tests that document behavior
+- Cover happy paths, edge cases, and error conditions
+- Use descriptive test names and clear assertions
+- Minimize test coupling and maximize maintainability
+- Balance thoroughness with practicality
+
+## Test Types You Generate
+- **Unit Tests**: Individual function/method testing
+- **Integration Tests**: Component interaction testing
+- **Edge Case Tests**: Boundary and error condition testing
+- **Performance Tests**: Basic performance validation
+
+## Code Analysis Skills
+- Identify testable units and their dependencies
+- Recognize complex logic that needs thorough testing
+- Spot potential failure points and edge cases
+- Understand mocking requirements for external dependencies`,
+  
+  instructionsPrompt: `Generate comprehensive tests for the provided code. Your process:
+
+1. **Analyze the codebase** using file-picker if needed to understand structure
+2. **Read target files** to understand functionality and dependencies
+3. **Identify test scenarios** including:
+   - Happy path cases
+   - Edge cases and boundary conditions
+   - Error handling scenarios
+   - Integration points
+4. **Generate test files** with:
+   - Proper test framework setup
+   - Descriptive test names
+   - Comprehensive assertions
+   - Appropriate mocks/stubs
+5. **Run tests** to ensure they work
+6. **Provide recommendations** for testing strategy improvements
+
+Focus on creating maintainable, readable tests that serve as documentation.`,
+  
+  handleSteps: function* ({ agentState, prompt, params }) {
+    // Step 1: Understand the codebase structure
+    yield {
+      toolName: 'spawn_agents',
+      args: {
+        agents: [{
+          agent_type: 'file-picker',
+          prompt: `Find files related to: ${prompt}. Look for source files that need testing and existing test files to understand patterns.`
+        }]
+      }
+    }
+    
+    // Step 2: Let the model analyze and generate tests
+    yield 'STEP_ALL'
+  }
+}
+
+export default config