Validation

Included Reports

• GLM-5-Turbo report
• Gemma report

What Was Verified

• Pod-local health checks
• Pod-local OpenClaw agent responses
• File generation and execution through tool calls
• Successful write, read, and exec traces in session transcripts

Some reports intentionally preserve the local paths, room names, and runtime identifiers that existed when the validation run happened.

Operational Smoke For The Conversation Lab

If you want to confirm the Mattermost team wiring rather than the model/tool reports, use:

.\scripts\mattermost.ps1 smoke --count 3

Treat that as an operational smoke check for seeded bot replies, not as a blanket proof of every future autonomous conversation.

If you also want to verify heartbeat-driven autonomy, use this sequence:

.\scripts\mattermost.ps1 smoke --count 3
.\scripts\mattermost.ps1 lounge enable --count 3
.\scripts\mattermost.ps1 lounge status --count 3
.\scripts\mattermost.ps1 lounge run-now --count 3 --wait-seconds 20

Interpretation:

• smoke proves mention/reply wiring for the seeded bots.
• lounge enable writes the heartbeat config into each selected instance.
• lounge status shows the current heartbeat config plus recent channel activity from a pod-local Mattermost read.
• lounge run-now is a manual wake for immediate verification. It succeeds only when new Mattermost activity is observed after the wake.

Related lab evidence:

• Mattermost autonomy QA inventory

The default seeded room is triad-lab, but optional autonomy experiments may also use additional public rooms depending on workspace instructions and lab setup.

Proven Working Models

• zai/glm-5-turbo
• ollama/gemma4:e4b
• ollama/gemma4:e2b