Evaluation Record Template #
The shape of an evaluation record: one item page per model or product, and one revision page per evaluation beneath it. Copy the skeletons below.
Where it goes #
| Hardware | Software | |
|---|---|---|
| Item | In service: the model’s page, content/docs/platforms/hardware/<group>/<model>/_index.md. A candidate: content/docs/platforms/evaluations/hardware/<model>/_index.md | content/docs/platforms/evaluations/software/<layer>/<product>/_index.md |
| Revision | …/<model>/<YYYY-MM>-hw-<revision>.md, e.g. 2026-10-hw-1-20.md | …/<product>/<line>.md, e.g. 26-7.md |
| Layer | network or compute, as on
Hardware | network, management-plane or tenant-compute, as on
Implementation & Tooling, or tooling for build and automation tools |
| Item title | The model, e.g. “TP-Link SG2218” | The product, e.g. “OPNsense” |
| Revision title | “2026-10 · hardware 1.20” | “26.7” |
| Weight | The first revision is 1000, and each new one is one lower, so the newest sorts first | the same |
| Index | In service: already listed on Hardware; update its Evaluation section. A candidate: add it to Evaluations | Add the item to its layer page, and to Evaluations |
Add bookCollapseSection: true to every item page, so its revisions stay folded in the sidebar.
For hardware in service, the item page already exists: it is the model’s page under Hardware. Its Evaluation section carries the Role and Current rows and the Revisions table from the skeleton below, and its revision pages go beneath it.
Item page skeleton #
---
title: "<Model or product>"
bookCollapseSection: true
---
# <Model or product>
| | |
|---|---|
| **Role** | <the role it is evaluated for, e.g. core router> |
| **Current** | {{< status-badge "complete" "Approved" >}} <revision or line> |
| **Selected on** | [<platform page>](/docs/platforms/...) |
| **In the catalog** | [Software Catalog](/docs/platforms/software-catalog/#...) *(software only)* |
<One paragraph: what it is and why it was evaluated.>
## Revisions
| Revision | Date | Verdict | Notes |
|---|---|---|---|
| [<revision>](<slug>/) | YYYY-MM-DD | Approved | <one line> |
Revision page skeleton #
---
title: "<revision>"
weight: <1000 for the first revision; one lower for each after>
---
# <Model or product>: <revision>
| | |
|---|---|
| **Verdict** | {{< status-badge "evaluating" "Under evaluation" >}} |
| **Date** | YYYY-MM-DD (evaluation started) · YYYY-MM-DD (verdict) |
| **Role** | <role> |
| **Tested with** | <hardware: the OS and driver · software: the hardware and the bundled components the role uses> |
| **Environment** | <bench / throwaway / on site, and what that means for the evidence> |
| **Supersedes** | <previous revision, or None> |
## Criteria
One row per criterion on the [hardware](/docs/policies/lifecycle-management/evaluation/hardware-evaluation/#criteria) or
[software](/docs/policies/lifecycle-management/evaluation/software-evaluation/#criteria) page, in order, including the Preferred ones.
| # | Criterion | Result | How it was checked | What was seen |
|---|---|---|---|---|
| H1 / S1 | <criterion> | Passed / Failed / Not tested / Not applicable (why) | <what was done> | <measurements, log lines, links> |
## Conditions
<Approved with conditions only: each Required criterion missed or not tested, and what would clear it. Otherwise "None".>
## Findings
<Anything learned that isn't a criterion result, and where it went: a risk, an incident, a follow-up.>
## Addenda
<Dated, appended only: patch releases re-checked against a criterion, and why.>