aceteam-nodes 0.6.0

AceTeam workflow nodes for local execution

aceteam-nodes

AceTeam workflow node types for local execution. Use the Ace CLI to run workflows; this package provides the Python runtime and optional tooling such as browser profile setup for Playwright-based nodes.

Install

pip install aceteam-nodes

Quick Start

Run workflows with the Ace CLI (ace workflow run). This repository ships node implementations and a small helper for Chromium profile setup used by BrowserFetch:

# After: uv sync --group playwright && playwright install chromium
ace-browser-setup
# equivalent: python -m aceteam_nodes  or  aceteam-nodes

Configuration

Create ~/.ace/config.yaml to configure API keys and defaults:

default_model: gpt-4o-mini

Set your API key via environment variable:

export OPENAI_API_KEY=sk-...
# or
export ANTHROPIC_API_KEY=sk-ant-...

How It Works

Workflow JSON ──> AceTeamWorkflow (load + validate)
                       │
                       ▼
              workflow-engine (DAG execution)
                       │
                       ▼
              AceTeamNode.run() per node
                  │         │
                  ▼         ▼
              litellm    httpx/jinja2
            (100+ LLMs)  (API calls)

1. A workflow JSON file defines nodes and their connections as a directed acyclic graph
2. AceTeamWorkflow validates the file and resolves node references
3. The aceteam-workflow-engine executes nodes in topological order
4. Each node type implements run() — LLM nodes call litellm, API nodes use httpx, etc.
5. Results flow through the graph until all outputs are produced

Available Nodes

Node	Description
LLM	AI text generation via litellm (100+ providers)
APICall	HTTP requests with Jinja templating
Equal, NotEqual, GreaterThan, LessThan, And, Or, Not	Comparison/logic operators

Development

# Setup
uv sync --extra dev

# Test
uv run pytest

# Lint & format
uv run ruff check && uv run ruff format

# Type check
uv run pyright

# Build
uv build

Related

• Ace CLI — TypeScript CLI that wraps this package for a streamlined user experience
• Workflow Engine — The DAG execution engine used under the hood

License

MIT