**Legend:** ▲ top/answer · ⚠ violation · ↑ buried conclusion · ⇄ wrong order · ⊕ gap · ✂ cut **Annotated text:** > This memorandum describes how project teams identify and deliver significant profit improvements. [⚠ topic label, not an answer — states what the document is about, not what it concludes] It is organized in six sections: [✂ table-of-contents framing substitutes for a key line] > > Background [⊕ not a claim — a label; unclear what it summarizes] > Principles of the project...
$npx -y skills add welltraum/minto --skill raw --agent claude-code
Installs into .claude/skills of the current project.
Are you the author of Raw?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/welltraum-raw-849e3dde)
**Legend:** ▲ top/answer · ⚠ violation · ↑ buried conclusion · ⇄ wrong order · ⊕ gap · ✂ cut
**Annotated text:**
> This memorandum describes how project teams identify and deliver significant profit improvements. [⚠ topic label, not an answer — states what the document is about, not what it concludes] It is organized in six sections: [✂ table-of-contents framing substitutes for a key line]
>
> Background [⊕ not a claim — a label; unclear what it summarizes]
> Principles of the project-team approach [⊠ kind unclear: mechanics or reasons?]
> Definition of project work [⚠ mixed kind — this is "what it is", sitting beside "how it works" and "why it succeeds"]
> Organization of the program [⚠ mechanics — same-kind as "principles", not separated]
> Distinctive benefits and results [↑ this is the one section that could carry the answer — payoff belongs at the top, not buried as item 5 of 6]
> Conditions required for success [⚠ different kind again — risk/prerequisite, not benefit or mechanism]
**Findings**
| Location | Violation | Fix |
|---|---|---|
| Opening sentence | Top is a topic label ("describes how..."), not an answer to a reader question | State the reader's question ("does this approach work, should we adopt it?") and answer it in one sentence |
| Six-item list overall | Six first-level groups — exceeds the 3–4 working-memory limit | Consolidate around 3–4 same-kind groups |
| Items 2, 3, 4 ("Principles", "Definition", "Organization") | Mixed kinds — one describes what the work is, two describe mechanics/structure, none states a reader payoff | Re-derive from the reader's question: if it's "why adopt this?", recast as reasons/benefits; if "how does it work?", recast as steps |
| Item 5 ("Distinctive benefits and results") | Buried conclusion — the one benefit-bearing heading sits mid-list instead of driving the top | Promote to the top-level answer; everything else becomes support for it |
| Item 6 ("Conditions required for success") | Different kind (prerequisites/risks) mixed in with structural and benefit headings | Separate into its own class, or fold under whichever top-level group it actually supports |
| No stated reader question anywhere | Cannot verify vertical or horizontal logic without knowing what question the memo closes | Supply the question explicitly before further structuring |
| "Background" | Gap-marker heading, not a claim — impossible to tell what it summarizes or whether it belongs at this level at all | Either state what it concludes, or fold it into the Situation/Complication of an SCQ intro rather than a first-level section |
**Score**
| Axis | Score | Note |
|---|---|---|
| Top | 0 | No answer, only a topic description |
| Same-kind groups | 0 | Six labels of at least three different kinds (structure, mechanism, definition, benefit, precondition) |
| Order | 0 | No stated or inferable ordering principle — reads as recall order |
| MECE | 1 | Some doubtful overlap (e.g., "Principles" vs. "Organization"); cannot fully assess without the body |
| Display | 1 | Headings are visible as a list, but the list is not a hierarchy — it's flat and unranked |
**Total: 2/10 — not structured yet**
**Three fixes that buy the most:**
1. Name the reader's question, then write the answer as a single sentence — replace the descriptive opening.
2. Promote "distinctive benefits and results" (or whatever it turns out to support) from item 5 to the top.
3. Collapse six headings into 3–4 same-kind groups once the kind the question demands is established.