Patterns to archive - August 2026

Across all five categories assessed (Requirements, Architecture, Design, Development, Operations — 65 patterns total), here are the ones that passed **4 or fewer** of the 11 Core Quality Principles:

| Pattern | Passes | Key issue |
|---|---|---|
| [use-energy-efficient-hardware](https://patterns.greensoftware.foundation/architecture/technology-selection/use-energy-efficient-hardware) | 3 | Vendor marketing dressed as a pattern (links to a 2022 Ampere blog post, title borrows Ampere's own jargon); duplicates its own sibling [evaluate-other-cpu-architectures](https://patterns.greensoftware.foundation/architecture/technology-selection/evaluate-other-cpu-architectures). Recommend merging in, not fixing standalone. |
| [optimize-impact-on-customer-equipment](https://patterns.greensoftware.foundation/requirements/optimize-impact-on-customer-equipment) | 4 | Filed as Requirements but its Solution reads like a developer task list ("implement backwards-compatible code") — exactly the anti-pattern the skill's own guidance names. Needs reframing from the scoping angle, not a line-edit. |
| [evaluate-using-a-service-mesh](https://patterns.greensoftware.foundation/architecture/system-topology/evaluate-using-a-service-mesh) | 4 | Assumption ("no compliance requirements") is near-verbatim the skill's own textbook bad example. Same author (`yelghali`) repeats this exact template in [evaluate-whether-to-use-TLS-termination](https://patterns.greensoftware.foundation/development/evaluate-whether-to-use-TLS-termination) — worth flagging to that contributor specifically. |
| [reduce-network-traversal-between-VMs](https://patterns.greensoftware.foundation/architecture/system-topology/reduce-network-traversal-between-VMs) | 4 | Solution is one sentence + three bare vendor URLs standing in for actual guidance; no References section. Its one Consideration (multi-AZ vs. efficiency trade-off) is genuinely good and worth keeping. |
| [energy-efficent-framework](https://patterns.greensoftware.foundation/architecture/technology-selection/energy-efficent-framework) | 4 | Literal stub: Assumptions and Considerations both read "None." Needs a near-total rewrite. |
| [energy-efficent-models](https://patterns.greensoftware.foundation/architecture/technology-selection/energy-efficent-models) | 4 | Same "None/None" stub problem, plus names dated 2020-era models (DistilBERT, GPT-J) instead of durable technique categories. Very likely superseded by [right-sized-energy-efficient-ai-models](https://patterns.greensoftware.foundation/development/right-sized-energy-efficient-ai-models) (9/11) — recommend merging into that instead of revising. |
| [minify-web-assets](https://patterns.greensoftware.foundation/development/web-performance/minify-web-assets) | 4 | Third "None/None" stub. Doesn't note that most modern build tools (Vite, webpack, Next.js) already minify by default — a real gap given how dated that omission makes it feel. Overlaps with [remove-unused-css](https://patterns.greensoftware.foundation/development/media-and-code-efficiency/remove-unused-css). |

**Common thread:** all seven are missing Cost Impact; five of seven fail on testable Assumptions (vague hedge, tautology, or literal "None").

**Six of the seven are in Architecture** — that category's `technology-selection` folder in particular (3 of its 5 patterns land here) is the weakest spot found across the whole review. Operations and Design had none at this threshold, despite Operations carrying the worst *duplication* problem — its individual patterns were generally thin/incomplete rather than fundamentally broken on this many fronts at once.

Common thread across all seven: every one of them fails **Principle 5 (Complete Required Sections)** and **Principle 10 (Trade-off Visibility)** — all seven are missing Cost Impact — plus most also fail **Principle 8 (Testable Assumptions)**. Two of them (`energy-efficent-framework`, `energy-efficent-models`) are the literal "None"/"None" stub patterns flagged earlier. `use-energy-efficient-hardware` is the single weakest pattern found in this entire five-category review.

Start with the seven pattern links in the issue and review the Core Quality Principles, especially the required sections, Cost Impact, and testable Assumptions. Check the cited duplicate and superseding patterns before deciding whether each item should be archived, merged, or substantially rewritten. Done means the affected patterns have an agreed disposition and complete, non-placeholder guidance.

Written by the indexing model from the issue text.