{"id":46410013,"url":"https://github.com/aiagentflow/aiagentflow","last_synced_at":"2026-05-23T02:06:50.116Z","repository":{"id":341106153,"uuid":"1168930510","full_name":"aiagentflow/aiagentflow","owner":"aiagentflow","description":"A local-first, CLI-driven multi-agent AI software engineering workflow orchestrator with feed specs, PRDs, and guidelines to auto-generate implementation plans and code.","archived":false,"fork":false,"pushed_at":"2026-04-11T14:00:31.000Z","size":257,"stargazers_count":36,"open_issues_count":1,"forks_count":2,"subscribers_count":2,"default_branch":"main","last_synced_at":"2026-04-11T15:24:13.384Z","etag":null,"topics":["ai","ai-agents","anthropic","automation","cli","code-generation","developer-tools","gemini","llm","multi-agent","nodejs","ollama","openai","typescript","workflow"],"latest_commit_sha":null,"homepage":"https://aiagentflow.dev","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/aiagentflow.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-28T00:41:22.000Z","updated_at":"2026-04-11T13:58:22.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/aiagentflow/aiagentflow","commit_stats":null,"previous_names":["aiagentflow/aiagentflow"],"tags_count":11,"template":false,"template_full_name":null,"purl":"pkg:github/aiagentflow/aiagentflow","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aiagentflow%2Faiagentflow","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aiagentflow%2Faiagentflow/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aiagentflow%2Faiagentflow/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aiagentflow%2Faiagentflow/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/aiagentflow","download_url":"https://codeload.github.com/aiagentflow/aiagentflow/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aiagentflow%2Faiagentflow/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31686210,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-11T13:07:20.380Z","status":"ssl_error","status_checked_at":"2026-04-11T13:06:47.903Z","response_time":54,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai","ai-agents","anthropic","automation","cli","code-generation","developer-tools","gemini","llm","multi-agent","nodejs","ollama","openai","typescript","workflow"],"created_at":"2026-03-05T13:08:50.894Z","updated_at":"2026-05-23T02:06:50.085Z","avatar_url":"https://github.com/aiagentflow.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# aiagentflow\n\nA local-first CLI that orchestrates multi-agent AI workflows for software development. Give it a task — or feed it your specs, PRDs, and guidelines — and it coordinates specialized agents to architect, code, review, test, and ship automatically.\n\n**No cloud dependency. Bring your own API keys. Your code stays on your machine.**\n\n[![npm version](https://img.shields.io/npm/v/@aiagentflow/cli)](https://www.npmjs.com/package/@aiagentflow/cli)\n[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE)\n[![Node.js](https://img.shields.io/badge/node-%3E%3D20-green)](https://nodejs.org)\n\n---\n\n## How It Works\n\n```\nTask → Architect → Coder → Reviewer → Tester → Fixer → Ship\n```\n\nEach stage uses a specialized AI agent with tuned prompts and parameters. The loop repeats until quality thresholds pass — like a small AI engineering team running on your machine.\n\n---\n\n## Install\n\n```bash\nnpm install -g @aiagentflow/cli\n```\n\nOr with pnpm:\n\n```bash\npnpm add -g @aiagentflow/cli\n```\n\n---\n\n## Quick Start\n\n```bash\n# 1. Initialize in your project\ncd /path/to/your/project\naiagentflow init\n\n# 2. Run a task\naiagentflow run \"Add a login form with email/password validation\"\n\n# 3. Or run autonomously (no approval prompts)\naiagentflow run \"Refactor the auth module\" --auto\n\n# 4. Feed context docs to agents\naiagentflow run \"Add auth\" --context docs/api-spec.md docs/security.md\n\n# 5. Generate a task list from specs, then batch-run\naiagentflow plan docs/prd.md -o tasks.txt\naiagentflow run --batch tasks.txt --auto\n```\n\nThe `init` wizard walks you through:\n1. Auto-detect your project (language, framework, test framework, package manager)\n2. Select your LLM providers (Anthropic, OpenAI, Groq, Gemini, OpenRouter, Ollama)\n3. Enter API keys\n4. Assign models per agent role\n5. Choose a workflow mode (fast, balanced, strict)\n6. Import existing docs (specs, requirements, guidelines) for auto-loading\n\nConfiguration is saved locally in `.aiagentflow/config.json`.\n\n---\n\n## Features\n\n- **Multi-agent pipeline** — 6 specialized agents, each with a distinct role\n- **Context-aware** — feed specs, PRDs, architecture docs, and guidelines to every agent\n- **Plan from docs** — generate batch-ready task lists from your existing documentation\n- **Local-first** — runs entirely on your machine, no code leaves your system\n- **Provider-agnostic** — Anthropic, OpenAI, Groq, Google Gemini, OpenRouter (100+ models), Ollama (local/free)\n- **Workflow modes** — fast, balanced, or strict presets for iterations, approval, and temperatures\n- **Smart detection** — auto-detects language, framework, test runner, and package manager\n- **Configurable** — tune models, temperature, and iteration limits per agent\n- **Git-native** — auto-creates branches, auto-commits on QA pass\n- **Human-in-the-loop** — approve or override at any stage, or go full auto\n- **QA policies** — configurable quality gates (max critical issues, test requirements)\n- **Batch mode** — process multiple tasks from a file\n- **Session persistence** — crash recovery with automatic session saving\n- **Token tracking** — monitor LLM usage per agent and per run\n- **Customizable prompts** — edit agent prompts in `.aiagentflow/prompts/`\n\n---\n\n## CLI Commands\n\n| Command | Description |\n|---------|-------------|\n| `aiagentflow init` | Interactive setup wizard |\n| `aiagentflow config` | View current configuration |\n| `aiagentflow doctor` | Health check — verify providers and setup |\n| `aiagentflow run \u003ctask\u003e` | Run a workflow for a task |\n| `aiagentflow run \u003ctask\u003e --auto` | Autonomous mode (no approval prompts) |\n| `aiagentflow run \u003ctask\u003e --dry-run` | Preview the plan without executing |\n| `aiagentflow run \u003ctask\u003e --context \u003cfiles...\u003e` | Run with reference documents |\n| `aiagentflow run --batch tasks.txt` | Process multiple tasks from a file |\n| `aiagentflow plan \u003cdocs...\u003e` | Generate a task list from documentation |\n| `aiagentflow plan \u003cdocs...\u003e -o tasks.txt` | Write task list to file (batch-ready) |\n| `aiagentflow resume` | Resume the last interrupted session |\n| `aiagentflow sessions` | List all saved sessions |\n\n---\n\n## Agent Roles\n\n| Agent | Role | What it does |\n|-------|------|-------------|\n| 🧠 Architect | Plan | Analyzes the task and creates an implementation plan |\n| 💻 Coder | Implement | Writes production-ready code based on the plan |\n| 🔍 Reviewer | Review | Reviews code for bugs, security, and quality |\n| 🧪 Tester | Test | Generates tests and runs them |\n| 🐛 Fixer | Fix | Addresses review comments and test failures |\n| ✅ Judge | QA | Final quality gate — pass or fail |\n\n---\n\n## Supported Providers\n\n| Provider | Type | Default Model | Setup |\n|----------|------|---------------|-------|\n| **Anthropic** | Cloud API | `claude-sonnet-4-20250514` | Requires API key |\n| **OpenAI** | Cloud API | `gpt-4o-mini` | Requires API key |\n| **Groq** | Cloud API | `llama-3.3-70b-versatile` | Requires API key (generous free tier) |\n| **Google Gemini** | Cloud API | `gemini-2.0-flash` | Requires API key |\n| **OpenRouter** | Cloud API | `meta-llama/llama-3.1-8b-instruct:free` | Requires API key (100+ models, many free) |\n| **Ollama** | Local | `llama3.2:latest` | Requires [Ollama](https://ollama.com) running locally |\n\nYou can mix providers — use cloud APIs for reasoning agents (architect, reviewer, judge) and local models for generation agents (coder, tester, fixer).\n\n### Using with Ollama (free, local)\n\n```bash\n# Install and start Ollama\nollama serve\n\n# Pull a model\nollama pull llama3.2\n\n# Initialize aiagentflow with Ollama\naiagentflow init\n# → Select \"ollama\" as provider\n# → Enter model name: llama3.2\n```\n\n---\n\n## Configuration\n\nAfter `aiagentflow init`, your project has:\n\n```\n.aiagentflow/\n├── config.json              # Main configuration\n├── prompts/                 # Customizable agent prompts\n│   ├── architect.md\n│   ├── coder.md\n│   ├── reviewer.md\n│   ├── tester.md\n│   ├── fixer.md\n│   └── judge.md\n├── policies/                # Quality standards\n│   └── coding-standards.md\n├── context/                 # Reference docs (auto-loaded into every run)\n│   ├── api-spec.md          # Example: your API specification\n│   └── requirements.md      # Example: your PRD or requirements\n└── sessions/                # Saved workflow sessions\n```\n\nEdit the prompt files to customize how each agent behaves. Edit `coding-standards.md` to set project-specific rules that all agents follow. Drop `.md` or `.txt` files into `context/` and they'll be automatically included as reference material for all agents.\n\n---\n\n## Context Documents\n\nAgents work best when they understand your project's requirements, API contracts, and standards. There are three ways to provide reference documents:\n\n**1. Auto-loaded (recommended)** — Drop files into `.aiagentflow/context/`:\n\n```bash\ncp docs/api-spec.md .aiagentflow/context/\ncp docs/security-guidelines.md .aiagentflow/context/\naiagentflow run \"Implement user registration\"\n# Both docs are automatically included in every agent's context\n```\n\n**2. Per-run via `--context` flag:**\n\n```bash\naiagentflow run \"Add OAuth support\" --context docs/oauth-spec.md docs/auth-arch.md\n```\n\n**3. During init** — The setup wizard asks if you have existing docs and copies them for you.\n\n### What to include\n\n| Document type | Example | Why it helps |\n|---------------|---------|-------------|\n| API specs | `api-spec.md` | Agents generate correct endpoints and contracts |\n| Requirements / PRDs | `requirements.md` | Architect plans match your actual requirements |\n| Security guidelines | `security.md` | Reviewer catches violations against your policies |\n| Architecture docs | `architecture.md` | Coder follows your patterns and conventions |\n| Development guidelines | `dev-guidelines.md` | All agents follow your team's standards |\n\n### Plan command\n\nTurn documentation into an actionable task list, then batch-run it:\n\n```bash\n# Generate tasks from a PRD\naiagentflow plan docs/prd.md -o tasks.txt\n\n# Review the generated tasks\ncat tasks.txt\n\n# Run them all\naiagentflow run --batch tasks.txt --auto --context docs/architecture.md\n```\n\n---\n\n## Project Structure\n\n```\nsrc/\n├── cli/            # CLI entry point and commands\n├── core/           # Config system, workflow engine, QA policies\n├── providers/      # LLM provider adapters (Anthropic, OpenAI, Groq, Gemini, OpenRouter, Ollama)\n├── agents/         # Agent implementations and prompt library\n├── git/            # Git operations wrapper\n├── prompts/        # Default prompt templates\n└── utils/          # Shared utilities (logger, fs, validation)\n```\n\n---\n\n## Development\n\n```bash\n# Clone and install\ngit clone https://github.com/aiagentflow/aiagentflow.git\ncd aiagentflow\npnpm install\n\n# Run in dev mode\npnpm dev run \"your task here\"\n\n# Type check\npnpm typecheck\n\n# Run tests\npnpm test\n\n# Lint \u0026 format\npnpm lint\npnpm format\n```\n\n---\n\n## Contributing\n\nContributions are welcome! Here's how to get started:\n\n1. **Fork** the repo and clone your fork\n2. **Create a branch** for your feature: `git checkout -b feature/your-feature`\n3. **Follow the coding standards:**\n   - Functions: `camelCase`, Classes: `PascalCase`, Files: `kebab-case`\n   - All public functions need JSDoc, types, and error handling\n   - Use custom `AppError` subclasses — never raw `throw new Error()`\n4. **Check your work:** `pnpm typecheck \u0026\u0026 pnpm lint \u0026\u0026 pnpm test`\n5. **Open a PR** against `main` with a description of what and why\n\n### Architecture rules\n\n- Dependency direction flows downward: `cli → core → utils → types`\n- Config types are inferred from Zod schemas, never manually defined\n- New providers only require one adapter file + registry entry\n\n---\n\n## Roadmap\n\n- [x] Project scaffolding, config system, LLM provider layer\n- [x] Workflow engine, agent implementations, Git integration\n- [x] QA policies, token tracking, session persistence\n- [x] Context documents — feed specs, PRDs, and guidelines to agents\n- [x] Plan command — generate task lists from documentation\n- [x] Multiple providers — Anthropic, OpenAI, Groq, Gemini, Ollama\n- [x] Project auto-detection — language, framework, test runner, package manager\n- [x] Auto-commit on QA pass\n- [x] Workflow mode presets — fast, balanced, strict\n- [ ] VSCode extension\n- [ ] Desktop GUI\n\n---\n\n## License\n\n[MIT](LICENSE)\n\n---\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://aiagentflow.dev\"\u003eaiagentflow.dev\u003c/a\u003e\n\u003c/p\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faiagentflow%2Faiagentflow","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faiagentflow%2Faiagentflow","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faiagentflow%2Faiagentflow/lists"}