bokmaal-proof — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited bokmaal-proof (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
LLMs and most language tools handle Norwegian bokmål grammar reasonably well. They miss the things native Norwegian readers actually notice: anglicisms ("skalere opp" instead of "vokse"), register drift (formal-to-casual mid-paragraph), translation-from-English sentence rhythm, and Språkrådet recommendations that no tool enforces. This skill fixes those.
Core principle: the test is "would a Norwegian reader feel this was written by a Norwegian", not "is this grammatically correct".
/bokmaal-proof or /bokmål-proof with Norwegian text or a path to a Norwegian document.Do NOT use for: nynorsk (this skill is bokmål-only; tell the user politely if they want nynorsk), creative fiction (the rules below are for technical and business writing), or quick spell-check (a normal spellchecker is faster).
Ask the user, in one batched question if any are unclear:
If the user does not answer, default to: profesjonelt-nøytral, fagpublikum, in-place rewrite.
Read the text six times, looking for one class of issue per pass. Do not try to do all six in one pass; you will miss things.
| Pass | Looking for | Examples |
|---|---|---|
| 1. Anglicisms | English words and English-shaped phrases that have a Norwegian equivalent | "skalere opp" -> "vokse"; "leverage" -> "utnytte" or "bruke"; "stakeholders" -> "interessenter" or "berørte parter"; "deliverable" -> "leveranse" |
| 2. Sentence rhythm | Sentences that follow English subject-verb-object-adverb patterns translated literally | English-rhythm: "Vi skal levere dette innen fredag." Norwegian-rhythm: "Dette leverer vi innen fredag." Norwegian tolerates topicalization much more than English. |
| 3. Register drift | Mixing formal and casual within the same paragraph | "Vi har gjennomført en analyse" + "stuff som funker" in same para = drift. Pick one register. |
| 4. Pronoun and tone | Inconsistent or inappropriate du/dere/De usage; passive overuse | Public-sector formality usually wants "du" today, not "De". Avoid passive when active is shorter. |
| 5. Språkrådet recommendations | Choices Språkrådet has explicitly recommended | "e-post" not "epost"; "nettside" not "webside"; "applikasjon" only when "app" or "program" does not fit; "kunstig intelligens" or "KI", not "AI" in formal Norwegian writing. |
| 6. Compound words | Wrongly split compounds (særskriving), the most common Norwegian writing error | "lammekoteletter" (right) not "lamme koteletter" (wrong); "barnehage" not "barne hage". This single error makes text look amateur faster than any other. |
Two output modes depending on user preference:
In-place rewrite mode: Return the cleaned text with no annotations, plus a short summary at the end:
## Endringer
- N anglicismer erstattet
- N setninger omformulert for bedre rytme
- N særskrivinger rettet
- N Språkrådet-anbefalinger fulgt
- N andre justeringer
## Største mønster
[1 til 2 setninger om det viktigste stilistiske mønsteret som gikk igjen]Comments-only mode: Return the original text annotated with inline suggestions in the form [opprinnelig -> forslag (begrunnelse)]. Add the same summary block at the end.
Offer one follow-up: "Vil du at jeg skal kjøre samme gjennomgang på resten av dokumentet, eller justere registeret?"
| English-influenced | Better bokmål |
|---|---|
| skalere opp | vokse, utvide |
| leverage | utnytte, bruke, dra nytte av |
| stakeholders | interessenter, berørte parter |
| deliverable | leveranse |
| timeline | tidsplan, framdriftsplan |
| roadmap | veikart, plan |
| insights | innsikt (singular i bokmål) |
| onboarding | innfasing, opplæring |
| commitment | forpliktelse, innsats |
| alignment | samstemthet, enighet |
| use case | bruksområde, brukstilfelle |
| pain point | problemområde, smertepunkt |
| best practice | beste praksis (akseptert), eller "anbefalt framgangsmåte" |
| AI | KI eller "kunstig intelligens" i formelle tekster |
These are starting points, not absolutes. Some anglicisms are now standard ("software" is fine in tech contexts; "programvare" is more formal). Use judgment.
| Failure | Fix |
|---|---|
| Caught grammar but missed anglicisms | Run the six passes separately. The anglicism pass is the highest-value one. |
| Particularly: missed særskriving | Compound-word splits are the single most damaging error. Always run pass 6. |
| Forced the user out of an acceptable anglicism | Some are now standard. Do not over-correct. Flag, do not always replace. |
| Mixed registers in the rewrite | Pick the user's chosen register and hold it. |
| Used nynorsk forms by mistake | Bokmål only. Common slip: "ikkje" -> "ikke", "skole" not "skule". |
| Replaced "AI" with "KI" in informal text | "AI" is fine informally. Only switch to "KI" in formal or public-sector contexts. |
| Returned everything in English | Output Norwegian. The summary block is also in Norwegian. |
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.