# Human-AI Work Toolkit

> Practical exercises for critical judgment, human-AI workflows and reusable agent guidance.

Start with the setup. Retrieve only the selected session, activity and its relevant material; preserve human decisions, prerequisites and reveal checkpoints.

## Start

- [Assistant setup](https://humanaiwork.org/ai/setup.md)
- [Method and delivery profiles](https://humanaiwork.org/ai/method.md)
- [Materials and worksheet prompts](https://humanaiwork.org/ai/materials.md)
- [Website](https://humanaiwork.org/)

## Sessions

- [Judgment Sampler](https://humanaiwork.org/ai/sessions/judgment-sampler.md): About 45 min; Sharpen one judgment rule by examining an AI response.
- [Agency Clinic](https://humanaiwork.org/ai/sessions/agency-clinic.md): About 1 hour; Map who decides at 3 moments in a workflow.
- [Human-AI Workflow Studio](https://humanaiwork.org/ai/sessions/human-ai-workflow-studio.md): Half day; Define how people and AI work together, with clear roles and review points.
- [Skill-to-Evidence Lab](https://humanaiwork.org/ai/sessions/skill-to-evidence-lab.md): About 2½ hours; Build and test AI guidance, then choose what to change.
- [Core Practices for Human-AI Work](https://humanaiwork.org/ai/sessions/core-practices-human-ai-work.md): Full day; Work through all 5 practices, from human judgment to a shared workflow and tested AI guidance.

## Offline reference

- [Complete core context, including checkpoint-separated references](https://humanaiwork.org/ai/toolkit-context.md)