Tutorial: grade sticky_elephant with UHBS (postgres)

Upstream: betheroot/sticky_elephant · UHBS: 4.2.2

docker network create uhbs-lab 2>/dev/null || true
docker build -f .local/labs/sticky_elephant/Dockerfile.lab -t sticky_elephant:uhbs-lab .local/labs/sticky_elephant
docker run -d --name sticky_elephant-lab --network uhbs-lab -p 127.0.0.1:15433:<container-port> sticky_elephant:uhbs-lab

UHBS_QUICK=1 UHBS_AIRGAP_ATTESTED=1 uhbs-lab \
  --inventory .local/sticky_elephant-inventory.yaml --target sticky_elephant-postgres \
  --tps docs/conformance/labs/sticky_elephant/<tps_quick>.yaml --protocol postgres \
  --phases profile,static,sandbox,dynamic,score --quick --skip-sast-tools \
  --out docs/conformance/reports/sticky_elephant/postgres/quick

Published: quick 40.35 / F, full 38.06 / F.

What you get from this lab

This tutorial reproduces the published UHBS evaluation proof for analysts who need to verify numbers, not trust a screenshot. After a successful run you should have SCORECARD.txt, report.json, and optional telemetry under .local/labs/sticky_elephant-telemetry.

How CTI / blue team should use the artifacts

Open the full SCORECARD first (authoritative). Treat quick as a smoke check. Read modules A–F with READING-UHQS.md. Confirm δ_C before citing UHQS externally. Wire your own log shipping; Module C is harness visibility, not SIEM coverage.

Trust limits

UHBS 4.2.2 evaluation proof is informative — not a certification or endorsement. Re-run after upstream or TPS changes; do not invent scores without regenerating artifacts.