Files
openclaw/qa/scenarios/config/system-agent-ring-zero-setup.yaml

38 lines
1.9 KiB
YAML

title: OpenClaw ring-zero setup
scenario:
id: system-agent-ring-zero-setup
surface: config
coverage:
primary:
- cli.service-auth-wiring-create
- cli.service-auth-wiring-system-agent-setup
secondary:
- channels.setup-onboarding-flows-system-agent-setup
- discord.token-and-application-id-configuration
- channels.setup-onboarding-flows-create
objective: Verify the package-installed OpenClaw inference gate and supporting one-shot setup operations, without claiming the interactive agent/tool/approval flow.
successCriteria:
- OpenClaw fails closed in an empty state dir and directs the user to inference onboarding.
- The packaged activation module tests a fake Claude inference backend before persisting its model, with remaining setup deferred to OpenClaw.
- A fuzzy packaged CLI request reaches the verified planner and resolves to a typed setup operation only after activation.
- The approved openclaw.setup operation independently writes the default workspace and model.
- The approved agents.create operation independently creates a non-main agent with its own workspace.
- Discord is enabled and configured through an env SecretRef without persisting the raw token.
- Config validation passes and audit entries exist for every applied write.
docsRefs:
- docs/cli/setup.md
- docs/channels/discord.md
- docs/help/testing.md
codeRefs:
- scripts/e2e/system-agent-first-run-docker.sh
- scripts/e2e/system-agent-first-run-spec.json
- test/e2e/qa-lab/runtime/system-agent-first-run-docker-client.ts
execution:
kind: script
path: test/e2e/qa-lab/runtime/docker-e2e-lane.ts
summary: Runs the packaged inference-gate and typed-operation Docker lane, with separate assertions for openclaw.setup and agents.create; interactive channel coverage remains secondary.
args:
- --lane
- system-agent-first-run