waxell 0.1.0

Waxell — build and run governed AI agents. Execution engine, authoring SDK, and the wax CLI.

waxell

Docs · Quickstart · CLI Reference · Observability · Coming from LangChain/CrewAI?

Build and run governed AI agents. One install gets you the durable execution engine (spawn/suspend, budget enforcement, governance hooks, lineage), the authoring decorators, the wax CLI, and observability:

pip install waxell

Using another agent framework and just want observability + governance on top? That's the standalone observe SDK:

pip install waxell-observe   # LangChain, OpenAI Agents, CrewAI, … — no runtime needed

What you get

Piece	What it is	How you touch it	Docs
waxell	The library — author + run agents	from waxell import agent, workflow, tool	Runtime quickstart
wax	The CLI — push, observe, govern from the terminal	wax push, wax runs, wax traces, …	CLI reference
waxell.observe	Observability SDK (the waxell-observe package)	mostly automatic	Observe overview

Quick start

from waxell import agent, workflow

@agent(name="lead_research", description="Research a lead and draft outreach")
class LeadResearch:
    @workflow("draft_email")
    async def draft(self, ctx, domain: str):
        company = await ctx.tool("fetch_company", domain=domain)
        email = await ctx.llm.generate(
            prompt=f"Write a 3-sentence intro for {company['name']}",
            task="outreach_email",
        )
        return {"email": email, "company": company["name"]}

Then push and observe it with the CLI (full command reference):

wax login                 # authenticate (or: wax config init)
wax push ./my-agents      # push agents, tools, workflows, and skills
wax runs                  # browse execution runs
wax traces                # explore distributed traces

Full walkthrough: runtime quickstart · agent/workflow/tool spec reference: SDK docs

What's in the box

• Execution engine — ExecutionContext, WorkflowEnvelope, durable Step, checkpoint/resume, ctx.spawn / ctx.spawn_many, tree-budget enforcement.

• Governance — pluggable GovernanceHook, in-runtime policy bridge, AXID mint/verify, lineage.

• Framework adapters — run governed LangGraph, CrewAI, Pydantic AI, AutoGen, and Claude Agent SDK agents (migration guides). Each adapter's framework dependency is optional:

  pip install "waxell[langgraph]"
  pip install "waxell[all]"     # every adapter
  

• Authoring — agent, workflow, router, tool, policy decorators, spec types, memory helpers, and typed Context objects for IDE autocomplete.

• Observability — every run traced; cost, tokens, and policy decisions in the Waxell dashboard via waxell.observe.

Import aliases: from waxell import agent is the documented path; from waxell_runtime import agent works too and is kept forever. The internal waxell_sdk package bundled in this wheel is an implementation detail — never import it directly.

Development (monorepo)

# Editable install keeps protected modules as readable .py (no Cython)
pip install -e runtime/waxell-runtime

# The internal waxell_sdk is synced into src/ at build/install time from
# ../../sdk/waxell-sdk — that synced copy is gitignored. Source of truth is
# agentforge/sdk/waxell-sdk/.

cd runtime/waxell-runtime && hatch run types:check
pytest runtime/waxell-runtime/tests/ -v

License

waxell is distributed under the terms of the MIT license.