✅ UAM Rating Example (v2)

This page shows a concrete, vendor‑neutral example of how to rate a platform against the Universal Activity Model (UAM) v2 and compute the Overall Activity Maturity (0–4).

How to use this

Review the seven UAM phases and their level definitions in the UAM Baseline (v2).
For your platform, assign a 0–4 level per phase based on evidence (screenshots, exports, policies, logs).
Apply gates (if missing governed retrieval or exportable evidence packs, cap at ≤2; see baseline notes).
Compute Overall Activity Maturity = round(nearest) of the average across phases after applying any caps.
Save a machine‑readable record under _data/ratings/uam/*.yaml so pages can render tables/cards consistently.

Tip: Use the YAML file in the Download section below as a template.

Example: “AcmeOps Platform” — UAM v2 rating

Overall Activity Maturity (0–4): 3 — Governed
Computed from per‑phase levels with no gating caps triggered.

Per‑phase scoring (with evidence notes)

UAM Phase	Level (0–4)	Evidence you should attach (examples)	Notes
1. Capture & Sessionization	4	Transcript export showing durable `base_id`, monotonic timestamps	Sessions persist across channels (chat + tool); correct `actor/channel/doc_type`.
2. Evidence Policy & Verification	3	Allow‑list config (domains/params), verify logs, TTL‑dedupe metrics	Implements allow‑lists + hashing; TTL dedupe partial (12h) → not full 4.
3. Sanitization & Redaction	3	Redaction test logs; commit/path normalization examples	Secrets/PII masked; path + commit normalization in place.
4. Enrichment & Linking	3	Hop graph export; manifests of links (commits/PRs/paths/tickets)	Hops & blast‑radius narrative generated for incidents.
5. Indexing & Search	3	Vector + metadata search traces; rerank proof	Combined search + rerank improves textual match quality.
6. Replay & Summarization	4	Replay Markdown + TL;DR with citations to verified evidence	Role‑labeled, noise‑pruned replay; TL;DR cites verified artifacts.
7. Governance, Export & Operability	3	RBAC/tenancy tests; Markdown/JSON exports; health/metrics dashboard	Governed retrieval enforced; exports present; SLOs tracked.

Average (unrounded) = (4+3+3+3+3+4+3) / 7 = 3.29 → rounded = 3.

Gates checked for this example

Exportable evidence packs? ✅ Present → no cap
Governed retrieval (tenant/user/visibility)? ✅ Present → no cap
Verification & dedupe discipline? ⚠️ Partial → kept level 3 for F2; still no global cap

JSON/YAML record (machine‑readable)

Use this structure under _data/ratings/uam/ to drive badges/cards in pages.

platform: "AcmeOps Platform"
version: "2025.10"
baseline: "UAM v2"
atlas_token_activities: "Y/M"   # Optional: if you also track Atlas token view
levels:
  capture_sessionization: 4
  evidence_policy_verification: 3
  sanitization_redaction: 3
  enrichment_linking: 3
  index_search: 3
  replay_summarization: 4
  governance_export_operability: 3
gates:
  evidence_packs_exportable: true
  governed_retrieval: true
  verification_dedupe: "partial"
composite:
  average_unrounded: 3.2857
  rounded: 3
  cap_applied: null
  overall: 3
evidence_links:
  transcript_export: "https://example.com/permalink/timeline/123"
  evidence_pack: "s3://acme-evidence/packs/2025-10-10/evidence.json"
  governance_tests: "https://example.com/run/ci/governance-tests-#9876"
notes: |
  TTL‑dedupe runs at 12h; move to 72h across sources to qualify for Level 4 on F2.

Download

Example page (this file): You can copy it to docs/capabilities/baselines/uam/rating-example.md.
Example data record: Place the YAML under docs/_data/ratings/uam/acme_uam_rating_v2.yaml.

Files:

Update steps (copy/paste)

Drop the YAML into _data/ratings/uam/ and change the values to your platform.
Copy this Markdown into docs/capabilities/baselines/uam/rating-example.md.
Link to the baseline: [UAM Baseline (v2)](/devops-sre-ai-atlas-2025/capabilities/baselines/uam/).
Commit & push; your sidebar will show UAM Rating Example (v2) under UAM.
For each vendor page, include a small badge or table that reads from _data/ratings/uam/*.yaml.