Next-level Agentic Orchestration Platform

id: go-code-analyzer
version: 1.0.0
description: AI-powered Go code analysis workflow

tasks:
  - id: parallel_analysis
    type: parallel
    description: Run performance and best practices analysis
    strategy: wait_all
    tasks:
      - id: performance_analysis
        type: basic
        agent: analyzer
        action: analyze
        with:
          file_path: "{{ .workflow.input.file_path }}"
          content: "{{ .workflow.input.file_content }}"

      - id: best_practices_analysis
        type: basic
        agent: analyzer
        action: analyze
        with:
          file_path: "{{ .workflow.input.file_path }}"
          content: "{{ .workflow.input.file_content }}"

resource: agent
id: analyzer
description: Specialized in Go code analysis for performance and best practices
version: 1.0.0

model: openai-gpt-4o-mini

instructions: |
  You are an expert Go developer specializing in performance optimization,
  best practices, concurrency, and memory management. Analyze code and
  provide actionable reports with specific, implementable suggestions.

tools:
  - write_file

actions:
  - id: analyze
    prompt: |-
      Analyze the following Go code file and save the review as a markdown file.
      Create a comprehensive markdown report with proper structure.
      Return a summary of your findings after saving the review file.

# Basic Task - Single operation execution
- id: analyze_code
  type: basic
  description: Analyze Go code for performance issues
  agent: analyzer
  with:
    file_path: "{{ .input.file_path }}"
    content: "{{ .input.file_content }}"

# Parallel Task - Concurrent execution
- id: multi_analysis
  type: parallel
  strategy: wait_all
  tasks:
    - id: performance_check
      type: basic
      # ... performance analysis
    - id: security_scan
      type: basic
      # ... security analysis

# Collection Task - Process arrays
- id: process_files
  type: collection
  items: "{{ .input.file_list }}"
  task:
    type: basic
    tool: file_processor
    with:
      file: "{{ .item }}"

import { z } from 'zod';

const inputSchema = z.object({
  name: z.string(),
  language: z.enum(['en', 'es', 'fr']).default('en')
});

export async function greeting(input: z.infer<typeof inputSchema>) {
  const { name, language } = inputSchema.parse(input);
  const greetings = {
    en: `Hello, ${name}!`,
    es: `¡Hola, ${name}!`,
    fr: `Bonjour, ${name}!`
  };

  return {
    message: greetings[language],
  };
}

# Global MCP server defined at workflow level
mcps:
  - id: context7
    transport: stdio
    command: "npx"
    args: ["-y", "@upstash/context7-mcp"]

# Agent that uses the global MCP server
agents:
  - id: docs-assistant
    model: openai-gpt-4o-mini

    instructions: |
      You are a documentation assistant. Use the Context7 MCP
      server to fetch up-to-date library documentation.

    actions:
      - id: get-docs
        prompt: |
          Find documentation for: {{ .input.library }}
          Topic: {{ .input.topic }}
          Use the Context7 MCP tools to retrieve the latest docs.

# Signal Task - Publish events
- id: publish_analysis_complete
  type: signal
  signal:
    id: "analysis.complete"
    payload:
      analysis_id: "{{ .workflow.id }}"
      file_path: "{{ .input.file_path }}"

# Wait Task - Subscribe to events
- id: wait_for_approval
  type: wait
  wait_for: "analysis.approved"
  timeout: "5m"
  condition: 'signal.payload.analysis_id == workflow.id'
  processor:
    tool: some_tool

# Event-triggered workflow
id: deployment-workflow
triggers:
  - type: signal
    name: "analysis.approved"

# Scheduled Workflow Configuration
id: daily-code-analysis
schedule:
  cron: "0 2 * * *"  # Daily at 2 AM
  timezone: "UTC"

# Batch processing with scheduling
tasks:
  - id: discover_repositories
    type: basic
    tool: git_scanner
    with:
      scan_path: "/repos"

  - id: analyze_all_repos
    type: collection
    items: "{{ .tasks.discover_repositories.output.repositories }}"
    concurrency: 5  # Process 5 repos in parallel
    task:
      type: basic
      agent: analyzer
      with:
        file_path: "{{ .item.path }}"
        content: "{{ .item.content }}"

# Memory Configuration
memory:
  - id: conversation_history
    type: token
    max_tokens: 4000
    key_template: "user:{{ .input.user_id }}:session:{{ .input.session_id }}"
    ttl: "24h"

# Using memory in agents
agents:
  - id: code_reviewer
    resource: agent
    memory:
      - id: conversation_history
        key: "user:{{ .input.user_id }}:session:{{ .input.session_id }}"
        mode: read-write
    model: openai-gpt-4o-mini
    instructions: |
      You are a code reviewer with access to conversation history.
      Remember previous analysis and user preferences.