agent-runtime/testcases/machine_pa_context.md
Nico 925fff731f v0.17.0: User expectation tracking, PA retry loop, machine state in PA context
- Memorizer tracks user_expectation (conversational/delegated/waiting_input/observing)
- Output node adjusts phrasing per expectation
- PA retry loop: reformulates job on expert failure (all retries exhausted or tool skip)
- Machine state in PA context: get_machine_summary includes current state, buttons, stored data
- Expert writes to machine state via update_machine + transition_machine
- Expanded baked schema coverage
- Awareness panel shows color-coded expectation state
- Dashboard and workspace component updates

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-03-30 19:03:07 +02:00

42 lines
1.3 KiB
Markdown

# Machine State → PA Context
Tests that PA reads machine state when routing, and experts can write back to machines.
Validates: enriched machine summary, update_machine, transition_machine.
## Setup
- clear history
## Steps
### 1. Create a machine
- send: create a navigation machine called "wizard" with initial state "start" and a second state called "details"
- expect_trace: has machine_created
### 2. PA sees machine in context
- send: what machines are active on my dashboard?
- expect_response: contains "wizard" or "start"
### 3. Expert stores data on machine
- send: use update_machine to store region=Bayern on the wizard machine
- expect_response: contains "Bayern" or "region" or "stored" or "updated"
### 4. PA sees stored data
- send: what data is stored in my wizard machine?
- expect_response: contains "Bayern" or "region"
### 5. Expert transitions machine to details
- send: use transition_machine to move wizard to details state
- expect_response: length > 5
### 6. PA sees updated state
- send: what state is the wizard in now?
- expect_response: contains "details"
### 7. Expert transitions back
- send: use transition_machine to move wizard back to start
- expect_response: length > 5
### 8. Final state check
- send: tell me the current wizard state and stored data
- expect_response: contains "start"