Player checklist

Owner: Platform team Scope: Every MVP release candidate must pass or document exceptions.

The player journey covers: clone → install → select model or fake demo → select scenario → speak/type → NPC responds → state evolves → finish → debrief → replay → confirm no cloud inference.

Automated checks (run in CI)

Run with the fake runtime — no model download required.

# From repo root (requires convsim-core and prompt-composer installed):
python -m pytest tests/acceptance/test_player_text_path.py -v

Test class	What it verifies	CI label
`TestScenarioSelection`	Scenario library returns at least one scenario; behavioral_interview is listed with title, summary, difficulty	`[scenario-lib]`
`TestSessionStart`	Session creation returns `sess-` ID; starting delivers NPC opening event with non-empty content	`[text-session]`
`TestPlayerTurn`	Text turn returns player_turn + npc_turn events; NPC has a valid emotion; two turns both succeed	`[text-session]`
`TestStateEvolution`	`state_delta` is present in NPC event; applied delta is returned; state carries into next turn	`[text-session]`
`TestSessionEnd`	Explicit end returns `Ended` + `player_exit`; further turns are rejected 409	`[text-session]`
`TestDebrief`	Debrief generated after session ends; debrief on active session rejected 409; idempotent	`[debrief]`
`TestSessionRetrieval`	Completed session retrievable by ID; two sessions both retrievable; export available after debrief	`[text-session]`
`TestNoCloudInference`	Full session with `LOCAL_MODE=True` completes without `NetworkBlockedError`; fake runtime in use	`[offline]`

All automated checks must pass before any manual step begins.

Manual checks (sign-off required)

Run on a clean profile (~/.convsim/ freshly created or deleted).

P-M1 — Clone and install

git clone https://github.com/outrightmental/ConversationSimulator.git
cd ConversationSimulator
./scripts/setup.sh

Setup completes without errors
Python venv present at services/convsim-core/.venv/
node_modules installed at repo root and workspaces

P-M2 — First run with fake demo (no model)

./scripts/dev.sh   # keep running
# open http://127.0.0.1:7354

Home screen loads; no console errors; no “Failed to fetch” in network tab
Service status indicators show correct state (fake runtime expected on fresh install)
Model manager screen shows no automatic download

P-M3 — Install or select real model (optional — requires adequate hardware)

Navigate to Settings → Models.

License acceptance dialog appears before download activates
Download completes with SHA-256 checksum verification
Model status changes to "loaded"
Record model filename and size in the smoke log

P-M4 — Select scenario

Navigate to Scenarios.

At least one official pack is listed
Expanding pack shows scenarios with title, description, difficulty tag
Player can select Job Interview Basics → Behavioral Interview

P-M5 — Speak or type

Start the selected scenario.

NPC opening line delivered within 10 seconds
Typing a player turn and submitting returns NPC response within 30 s (fake) / 60 s (real CPU)
Transcript updates after each turn

P-M6 — State meters (optional)

Enable Show state meters in session settings.

State variables visible during conversation
Values change after turns that trigger state deltas

P-M7 — Finish and debrief

End the session.

End session button works and confirms session closed
Debrief screen loads with rubric scores and (if session long enough) narrative
Export / Download debrief option is present

P-M8 — Replay

Return to the scenario library.

Session history or past sessions accessible
Previously played sessions retrievable by ID

P-M9 — No cloud inference confirmation

During or after a full session:

No outbound requests in browser network tab to external LLM APIs
LOCAL_MODE test passes in CI (automated gate above)

Sign-off

Item	Result	Notes
Automated suite	PASS / FAIL
P-M1 Clone + install	PASS / FAIL / SKIP
P-M2 Fake demo	PASS / FAIL / SKIP
P-M3 Real model	PASS / FAIL / SKIP	optional
P-M4 Scenario select	PASS / FAIL / SKIP
P-M5 Speak/type	PASS / FAIL / SKIP
P-M6 State meters	PASS / FAIL / SKIP	optional
P-M7 Debrief	PASS / FAIL / SKIP
P-M8 Replay	PASS / FAIL / SKIP
P-M9 No cloud	PASS / FAIL / SKIP

Release decision: All automated checks PASS + manual checks P-M1 through P-M9 PASS (or documented SKIP with justification) required before MVP tag.