AISBF Dashboard Operations Runbook

Dashboard operations

AISBF Dashboard Operations Runbook

A practical dashboard runbook for daily provider checks, rotation health, autoselect policy changes, cache safety, and user-token operations.

Start Free More tutorials

What this runbook is for

Use this when AISBF is already installed and you need a repeatable operator workflow. The goal is boring reliability: inspect the dashboard, make scoped changes, verify with an API call, and leave a trail for the next person.

Daily checks

Provider reachability, route inventory, cache status, and unusual user-token activity.

Change windows

Safe provider swaps, model additions, rotation edits, and autoselect policy changes.

Incident response

Move traffic away from a failing backend without redeploying application code.

AISBF dashboard overview used for operational checks.

Start at the overview: confirm the dashboard can see providers, rotations, cache, and account state before changing anything.

Daily dashboard check

Check provider status

Open Providers and confirm the active endpoints are the ones your apps are expected to use. Disable experiments that should not receive production traffic.

Inspect rotations

Open Rotations and make sure each production route has at least two healthy candidates unless it is intentionally pinned.

Review autoselect rules

Check that privacy, cost, and latency policies still point at the right rotations. Keep application-facing route names stable.

Verify from outside the dashboard

Use the same token and base URL your application uses. Dashboard green lights are helpful; a real API request is better.

Command-line verification

export AISBF_BASE="https://aisbf.example.com"
export AISBF_USER="alice"
export AISBF_TOKEN="user-scoped-token"

curl -fsS -H "Authorization: Bearer $AISBF_TOKEN"   "$AISBF_BASE/api/u/$AISBF_USER/models"

curl -fsS -H "Authorization: Bearer $AISBF_TOKEN"   -H "Content-Type: application/json"   "$AISBF_BASE/api/u/$AISBF_USER/chat/completions"   -d '{"model":"autoselect:support-default","messages":[{"role":"user","content":"health check"}]}'

Safe change checklist

Change	Do first	Verify
Add provider	Create it disabled or with low priority if supported.	Direct model call, then add to rotation.
Edit rotation	Keep old provider available until the new path passes.	Call the rotation route three times and inspect errors/latency.
Change autoselect	Document the policy reason: privacy, cost, latency, or capability.	Call the autoselect route and confirm selected backend in logs/dashboard.
Enable cache	Only cache idempotent or safe prompts.	Repeat the same request and check cache metrics.

Operator rule: applications should depend on stable AISBF route names, not individual provider names. Providers can churn; routes should not.

Try AISBF

AISBF is open source and also available as a hosted service. During the current testing period, hosted Pro is temporarily available as unlimited access for €2/month.

See the €2 Pro test plan View source code