seeragents 0.0.9

Seer - Multi-Agent System for Evaluating AI Agents

Seer (seeragents)

Seer is a workflow builder with fine-grained control for creating and executing automated workflows with integrated tools and services. Build complex automation workflows with visual editing, AI-assisted development, and seamless integrations (Google Workspace, GitHub, and more).

Note: Package name is seeragents on PyPI (name conflict), but CLI command is seer.

Quick Start

git clone <repo> && cd seer
uv run seer dev

That's it! No installation needed. Starts Docker services (Postgres, MLflow, backend), installs dependencies in containers, tails logs, waits for readiness, and opens the workflow builder in your browser.

Installation (Optional)

Only needed if you want to use seer directly without uv run:

CLI only (lightweight):

pip install "seeragents[cli]"  # or: uv pip install "seeragents[cli]"

Full installation:

pip install seeragents  # or: uv pip install seeragents

Local development:

git clone <repo> && cd seer
uv venv && source .venv/bin/activate
uv pip install -e ".[cli]"  # Install CLI only
rehash  # Refresh shell command cache (zsh) or restart terminal
seer dev  # Now you can use 'seer' directly

Configuration

Create a .env file (automatically loaded):

# Required for workflow execution and AI assistance
OPENAI_API_KEY=...
ANTHROPIC_API_KEY=...  # Alternative to OpenAI

# Integrations
TAVILY_API_KEY=...  # For web search tools

# OAuth Configuration (for cloud deployments)
GOOGLE_CLIENT_ID=...
GOOGLE_CLIENT_SECRET=...

# Optional: Persistence and monitoring
DATABASE_URL=...  # PostgreSQL for workflow persistence
MLFLOW_TRACKING_URI=...  # MLflow for execution tracking

Check: uv run seer config or seer config (if installed)

Usage

Development:

uv run seer dev  # Recommended: no installation needed
# or if installed:
seer dev

Configuration:

uv run seer config           # Show current configuration
uv run seer config --format json  # JSON output format

Data Export:

uv run seer export <thread-id>      # Export workflow execution results
uv run seer export <thread-id> --format markdown  # Export in markdown format

Development Workflow

What runs where:

• Local: CLI tool (seer command) - lightweight (click/rich only)
• Docker: Backend API, Postgres, MLflow - all dependencies installed here

Steps:

1. Run: uv run seer dev (no installation needed!)
2. Code changes hot-reload via volume mounts (uvicorn --reload)
3. Access workflow builder at: http://localhost:5173/workflows?backend=http://localhost:8000
4. View logs: docker compose logs -f
5. Stop: docker compose down

Services started:

• Backend API (port 8000): FastAPI server with workflow execution engine
• Postgres (port 5432): Workflow and user data persistence
• MLflow (port 5001): Execution tracking and observability

API Keys & Integrations

Feature	Required Keys
Workflow Execution	OPENAI_API_KEY or ANTHROPIC_API_KEY
AI Chat Assistant	OPENAI_API_KEY or ANTHROPIC_API_KEY
GitHub Integration	GITHUB_TOKEN, GITHUB_CLIENT_ID/SECRET
Google Workspace	GOOGLE_CLIENT_ID/SECRET
Web Search	TAVILY_API_KEY
Persistence	DATABASE_URL (PostgreSQL)
Monitoring	MLFLOW_TRACKING_URI
Cloud Auth	CLERK_JWKS_URL, CLERK_ISSUER

Supported Integrations:

• Google Workspace: Gmail, Google Drive, Google Sheets
• GitHub: Repositories, Issues, Pull Requests
• Web Tools: Search, content fetching
• Database: PostgreSQL with read/write controls

Missing keys? Seer prompts interactively and supports OAuth flows.

Key Features

🛠️ Visual Workflow Builder

• Drag-and-drop interface for creating automation workflows
• Node-based editor with custom blocks and integrations
• Real-time workflow validation and execution

🤖 AI-Assisted Development

• Chat interface for workflow design and debugging
• AI suggestions for workflow improvements
• Intelligent error handling and recovery

🔗 Rich Integrations

• Google Workspace: Gmail, Drive, Sheets with OAuth
• GitHub: Repository management, issues, PRs
• Web Tools: Search, content fetching, APIs
• Databases: PostgreSQL with approval-based write controls

⚡ Advanced Execution Engine

• Streaming execution with real-time updates
• Interrupt handling for human-in-the-loop workflows
• Persistent state management with PostgreSQL
• MLflow integration for observability

🔒 Enterprise-Ready

• Self-hosted or cloud deployment options
• OAuth-based authentication (Clerk integration)
• Role-based access control
• Audit trails and execution history