Skip to content

Open call

Bring us an agent workflow.

Share an agent, evaluation, tool, or difficult decision. Start with the smallest example we can inspect together.

01

Share an evaluation or toolkit

Share an evaluation, benchmark, policy check, or reusable toolkit. Tell us what it checks, what failure looks like, and where other people can inspect or run it.

Open a pack proposal
02

Recommend an agent tool

Recommend a tool you have used in a real agent workflow. Tell us which problem it solves, where it works well, and where you would not use it.

Open a tool proposal
03

Suggest an agent to test

Name an agent or system and the claim you want tested. We will follow the workflow, show the evidence, and explain what still depends on a person.

Request a teardown
04

Review a workflow with us

Bring one agent task, what a good result looks like, and the decision that currently feels hard to trust. We respond to concrete workflow reviews within a week.

Start a workflow review

What happens next

Start with one example we can inspect

The fastest way to begin is a small example. Give us something we can run, compare, or discuss rather than a broad introduction.

  1. Choose one task

    Pick a real task the agent performs and show us the current workflow.

  2. Describe a good result

    Tell us what should happen, what must never happen, and which decisions still need a person.

  3. Share the smallest example

    A link, screenshot, scorecard, pull request, or short problem description is enough to begin.

Not sure which path fits?

Open a general workflow review. Tell us what the agent does, what a good result looks like, and where the decision currently feels uncertain.

Start a conversation