# When reviewing code, always: - Check for security vulnerabilities - Verify error handling - Suggest tests for edge cases # When reviewing code, always: - Check for security vulnerabilities - Verify error handling - Suggest tests for edge cases

%@ %@

%@ active — injected into every conversation %@ aktiv — injisert i alle samtaler A footnote indicating that some skills are active.

%@ chars %@ tegn A label showing the number of characters in the skill's content. The argument is the number of characters in the content.

%@ min %@ min

%@ pt %@ pt A label showing the current size of the GUI text. The value is a slider that lets the user adjust the size, with a label showing the current size in points.

%@s %@s

(always used) (alltid brukt)

(optional) (valgfri)

1. Open Paperless-NGX → Settings → API Tokens 1. Åpne Paperless-NGX → Innstillinger → API-tokens A step in the process of getting a Paperless-NGX API token.

2. Create or copy your token 2. Opprett eller kopier tokenet ditt A step in the process of getting a Paperless-NGX API token.

3. Paste it above 3. Lim det inn ovenfor A step in the process of getting a Paperless-NGX API token.

A shortcut with command %@ already exists. En snarvei med kommandoen %@ finnes allerede. A warning message that appears when a user tries to import a shortcut that already exists in their settings. The argument is the command of the conflicting shortcut.

API keys and credentials are **not** included in the backup. You will need to re-enter them after restoring on a new machine. API-nøkler og legitimasjon er **ikke** inkludert i sikkerhetskopien. Du må taste dem inn på nytt etter gjenoppretting på en ny maskin. A warning message displayed in the settings tab of the app.

Access Token Access Token

Active Aktiv A toggle

Active skills are appended to the system prompt. Toggle them per-skill to control what the AI knows. Aktive ferdigheter legges til systemprompten. Slå dem av og på per ferdighet for å kontrollere hva KI-en vet. A description of how to toggle the active status of a skill.

Agent Skills Agent-ferdigheter

Allow Shell Command? Tillat skalkommando? A title for a modal that asks the user if they want to allow a shell command.

Allow the AI to run shell commands on your machine. Commands are executed via /bin/zsh. Enable approval mode to review each command before it runs. Gi KI-en tilgang til å kjøre skalkommandoer på maskinen din. Kommandoer kjøres via /bin/zsh. Aktiver godkjenningsmodus for å se gjennom hver kommando før den kjøres.

Anthropic Balance Anthropic Balance A label displayed above the user's balance on the Anthropic platform.

Auto-execute mode: commands run without approval. Use with caution. Auto-kjøremodus: kommandoer kjøres uten godkjenning. Bruk med forsiktighet.

Auto-sync can cause conflicts if running on multiple machines simultaneously. Auto-synkronisering kan forårsake konflikter hvis den kjøres på flere maskiner samtidig.

Automatically select relevant messages instead of sending all history. Reduces token usage for long conversations. Velg automatisk relevante meldinger i stedet for å sende all historikk. Reduserer tokenforbruk for lange samtaler.

Automatically summarize old portions of long conversations to save tokens and improve context efficiency. Sammenfatt automatisk eldre deler av lange samtaler for å spare tokens og forbedre konteksteffektiviteten.

Back Up Now Sikkerhetskopier nå A button that backs up the user's data.

Backup location: Sikkerhetskopiplassering: A label displayed under the backup location.

Bash Execution Bash Execution

Branch: %@ Branch: %@

Built with SwiftUI Built with SwiftUI A label that says that the app is built with SwiftUI.

COMMAND KOMMANDO A label displayed above the command that the AI wants to run.

Capabilities Egenskaper A label displayed under the list of capabilities of a model.

Changing these values affects how the AI generates responses. The defaults work well for most use cases. Å endre disse verdiene påvirker hvordan KI-en genererer svar. Standardverdiene fungerer bra for de fleste brukstilfeller.

Check your balance at: Sjekk saldoen din på: A label displayed under a link to check the user's balance.

Check your usage at: Sjekk forbruket ditt på: A label displayed under a link to OpenAI's usage page.

Clear All Slett alle A button to clear all email activity logs.

Click 'Add Folder' below or drag folders here from Finder Klikk «Legg til mappe» nedenfor eller dra mapper hit fra Finder

Command Kommando A label displayed above the command field.

Command History Kommandohistorikk A label displayed above the command history.

Content (Markdown) Innhold (Markdown) A

Controls which messages are written to ~/Library/Logs/oAI.log Styrer hvilke meldinger som skrives til ~/Library/Logs/oAI.log

Conversations Samtaler

Copied! Kopiert! A message displayed when a code block is successfully copied to the clipboard.

Cost Examples Kostnadseksempler A heading for the cost examples of a model.

Create a dedicated email account specifically for AI handling. Do NOT use your personal email address. Opprett en dedikert e-postkonto spesielt for KI-håndtering. IKKE bruk din personlige e-postadresse.

Default Prompt Standardprompt

Delete (%@) Slett (%@)

Description Beskrivelse A label

Each command will require your approval before running. Hver kommando krever din godkjenning før den kjøres.

Edit Rediger A button to edit a skill.

Email Activity Log E-postaktivitetslogg The title of the email activity log view.

Email Handler System Prompt Systemprompt for e-postbehandler

Email handler uses ONLY its own system prompt, completely isolated from your main chat settings. A custom prompt below will override the defaults. E-postbehandleren bruker KUN sin egen systemprompt, fullstendig isolert fra de viktigste chatinnstillingene dine. En egendefinert prompt nedenfor vil overstyre standardinnstillingene.

Email tasks have READ-ONLY access to MCP folders. The AI cannot write, delete, or modify files when processing emails. E-postoppgaver har SKRIVEBESKYTTET tilgang til MCP-mapper. KI-en kan ikke skrive, slette eller endre filer ved behandling av e-post.

Enable AI-powered semantic search using %@ embeddings. Cost: ~$0.02–0.15/1M tokens. Aktiver KI-drevet semantisk søk med %@-innbygginger. Kostnad: ~$0,02–0,15/1M tokens.

Enable email handler in Settings to start monitoring emails Aktiver e-postbehandler i Innstillinger for å starte overvåking av e-post A message displayed when the email handler is disabled.

Error: %@ Feil: %@ A label displaying an error message in the checkout view.

Example structure: Eksempelstruktur: A

Example: oai-bot-x7k2m9p3@gmail.com Example: oai-bot-x7k2m9p3@gmail.com

File Access Permissions Filtillatelser

Files Filer A label displayed above the list of files attached to a skill.

Google (Gemini embedding) Google (Gemini embedding)

Help Hjelp A heading for the help screen

How to get your API token: Slik får du API-tokenet ditt: A heading for a section that describes how to get your API token.

Initialize Repository Initialiser repositorium

Inject into system prompt for every conversation Injiser i systemprompten for hver samtale

Large files inflate the system prompt and may hit token limits. Store filer blåser opp systemprompten og kan nå token-grenser. A warning displayed when a user adds a large file to a skill.

Last sync: %@ Siste synkronisering: %@

Leave empty to use the default email handler system prompt. La stå tomt for å bruke standard systemprompt for e-postbehandler.

Lowercase letters, numbers, and hyphens only. No spaces. Kun små bokstaver, tall og bindestreker. Ingen mellomrom. A description of the format of a shortcut's command.

MCP gives the AI controlled access to read and optionally write files on your computer. The AI can search, read, and analyze files in allowed folders to help with coding, analysis, and other tasks. MCP gir KI-en kontrollert tilgang til å lese og eventuelt skrive filer på datamaskinen din. KI-en kan søke, lese og analysere filer i tillatte mapper for å hjelpe med koding, analyse og andre oppgaver.

Markdown instruction files injected into the system prompt. Compatible with SKILL.md format. Markdown-instruksjonsfiler som injiseres i systemprompten. Kompatibel med SKILL.md-format. A description of the format of agent skills.

Max Tokens: set to 0 to use model default. Higher values allow longer responses. Maks tokens: sett til 0 for å bruke modellstandard. Høyere verdier gir lengre svar.

Model Context Protocol Model Context Protocol

Model Info Modellinformasjon A title for a view that displays information about a model.

Multi-provider AI chat client AI-chatklient med støtte for flere tilbydere A description of oAI.

Name Navn A label

Never Aldri A label displayed in the settings view that indicates when the last backup was taken.

No credit data available Ingen kredittdata tilgjengelig A message displayed when there is no credit data available.

No email activity yet Ingen e-postaktivitet ennå A message displayed when the user has not yet received any email activity.

No files attached. Add JSON, YAML, CSV or TXT files to inject data into the system prompt alongside this skill. Ingen filer vedlagt. Legg til JSON, YAML, CSV eller TXT-filer for å injisere data i systemprompten sammen med denne ferdigheten. A message displayed when a user has not attached any files to a skill.

No model selected Ingen modell valgt A label displayed when no model is selected.

No models available Ingen modeller tilgjengelig

No results for "%@" Ingen resultater for «%@» A message displayed when there are no search results. The argument is the search term.

No shortcuts yet Ingen snarveier ennå A message displayed when a user has no shortcuts.

No shortcuts yet — click New Shortcut to create one. Ingen snarveier ennå — klikk på Ny snarvei for å opprette en. A message displayed when a user has no shortcuts.

No skills yet Ingen ferdigheter ennå A message displayed when the user has no skills.

No skills yet — click New Skill or Import to get started. Ingen ferdigheter ennå — klikk på Ny ferdighet eller Importer for å komme i gang. A message displayed when a user has not created any skills.

Ollama (Local) Ollama (Local) A label displayed above the credits information for the local Ollie.

Only emails with this text in the subject line will be processed. Example: "[OAIBOT] What's the weather?" Bare e-poster med denne teksten i emnefeltet behandles. Eksempel: «[OAIBOT] Hva er været?»

OpenAI (text-embedding-3-large) OpenAI (text-embedding-3-large)

OpenAI (text-embedding-3-small) OpenAI (text-embedding-3-small)

OpenAI Balance OpenAI Balance A label displayed above the OpenAI balance section.

OpenRouter (OpenAI large) OpenRouter (OpenAI large)

OpenRouter (OpenAI small) OpenRouter (OpenAI small)

OpenRouter (Qwen 8B) OpenRouter (Qwen 8B)

OpenRouter Credits OpenRouter Credits A heading for the user's OpenRouter credits.

Read access (always enabled) Lesetilgang (alltid aktivert)

Remote: %@ Remote: %@

Restore from File… Gjenopprett fra fil… A button that allows the user to restore their data from a file.

Running locally — no credits needed! Kjører lokalt — ingen kreditter nødvendig! A message displayed when using an on-device LLM like the one provided by the `.ollama` provider.

SKILL.md format — write instructions in plain Markdown. SKILL.md-format — skriv instruksjoner i vanlig Markdown.

SSH Key SSH Key

Security Recommendation Sikkerhetsanbefaling

Shell commands have full access to your system. Only approve commands you understand and trust. Skalkommandoer har full tilgang til systemet ditt. Godkjenn bare kommandoer du forstår og stoler på. A warning banner displayed in the Bash Approval Sheet.

Skills are markdown instruction files that teach the AI how to behave. Active skills are automatically injected into the system prompt. Ferdigheter er Markdown-instruksjonsfiler som lærer KI-en hvordan den skal oppføre seg. Aktive ferdigheter injiseres automatisk i systemprompten. A description of how skills are used by the AI.

Stream responses as they're generated. Disable for single, complete responses. Strøm svar etter hvert som de genereres. Deaktiver for enkle, fullstendige svar.

Sync Now Synkroniser nå

Sync conversations and settings across multiple machines using Git. Synkroniser samtaler og innstillinger på tvers av flere maskiner ved hjelp av Git.

Temperature: 0 = model default · 0.0–0.7 = focused · 0.8–2.0 = creative Temperatur: 0 = modellstandard · 0,0–0,7 = fokusert · 0,8–2,0 = kreativ

Template Mal A label displayed above the text editor for the shortcut's template.

Test Connection Test tilkobling

Text files are injected into the system prompt alongside the skill. Tekstfiler injiseres i systemprompten ved siden av ferdigheten. A description of how text files are injected into the system prompt.

The AI can read and search files in allowed folders KI-en kan lese og søke filer i tillatte mapper

The AI reads this content and decides when to apply it. Describe **what** the AI should do and **how** — be specific and concise. KI-en leser dette innholdet og bestemmer når det skal brukes. Beskriv **hva** KI-en skal gjøre og **hvordan** — vær spesifikk og konsis.

The AI wants to run the following command KI-en ønsker å kjøre følgende kommando A description of the command that the AI wants to run.

This default prompt is always included to ensure accurate, helpful responses. Denne standardprompten er alltid inkludert for å sikre nøyaktige, nyttige svar.

This will permanently delete all email activity logs. This action cannot be undone. Dette vil slette alle e-postaktivitetslogger permanent. Denne handlingen kan ikke angres. A message displayed in an alert when the user is about to clear all email logs.

Token Distribution Token-fordeling A label displayed under the token distribution bar.

Try adjusting your search or filters Prøv å justere søket eller filtrene A description of the error that occurs when no models match the user's search.

Type / in the input to see command suggestions Skriv / i inndata for å se kommandoforslag A description of how to use the keyboard shortcut.

Type a message or / for commands... Skriv en melding eller / for kommandoer... A placeholder text displayed in the input area when

Uncommitted changes: %@ Ulagrede endringer: %@

Update Available%@ Oppdatering tilgjengelig%@ A button that opens a website with information about a new version of oAI. The argument is the version number of the new version.

Use **{{input}}** in the template to insert whatever you type after the command. Bruk **{{input}}** i malen for å sette inn det du skriver etter kommandoen. A tooltip for the "lightbulb" icon in the shortcuts view

Use **{{input}}** to insert whatever you type after the command. Bruk **{{input}}** for å sette inn det du skriver etter kommandoen. A description of how to use the shortcut template.

Use @filename to attach files to your message Bruk @filnavn for å legge ved filer i meldingen A description of how to attach files to a message.

Username + Password Username + Password

Version %1$@ (%2$@) Version %1$@ (%2$@) A version number in the format "Version 1.0 (1)"

View Email Log Vis e-postlogg

When enabled, listing and searching skip gitignored files. Write operations always ignore .gitignore. Når aktivert, hoppes gitignorerte filer over ved listing og søk. Skriveoperasjoner ignorerer alltid .gitignore.

Working directory: Arbeidsmappe: A label displayed under the working directory of a pending shell command.

You can import any SKILL.md file from skill0.io or write your own. Du kan importere en hvilken som helst SKILL.md-fil fra skill0.io eller skrive din egen. A description of how to use skills.

[GNU Affero General Public License v3.0](https://www.gnu.org/licenses/agpl-3.0.html) [GNU Affero General Public License v3.0](https://www.gnu.org/licenses/agpl-3.0.html) A link to the GNU Affero General Public License.

^[%@ entry](inflect: true) ^[%@ oppføring](inflect: true) A link that shows the number of email logs. The number is pluralized based on the value of the `inflect: true` modifier.

active aktiv A label for an active skill.

messages meldinger

oAI oAI The name of the app.

tokens tokens

v%@ v%@ A label showing the current version of oAI.

~750 tokens ≈ 500 words. Online mode allows web search in responses. ~750 tokens ≈ 500 ord. Nettmodus tillater nettsøk i svar.

• • A bullet point.

• Add public key to your git provider • Legg til den offentlige nøkkelen hos git-leverandøren din

• No credentials needed in oAI • Ingen legitimasjon nødvendig i oAI

• Uses your system SSH keys (~/.ssh/id_ed25519) • Bruker systemets SSH-nøkler (~/.ssh/id_ed25519)

ℹ️ SSH Key Authentication ℹ️ SSH-nøkkelautentisering

↑↓ navigate ↩ open ↑↓ navigate ↩ open

⌘N New • ⌘M Model • ⌘S Save ⌘N New • ⌘M Model • ⌘S Save A hint that appears on macOS when using keyboard shortcuts.

⚠️ Custom prompt active — only this prompt will be sent to the model. ⚠️ Egendefinert prompt aktiv — bare denne prompten sendes til modellen.

⚠️ Many providers (GitHub) no longer support password authentication. Use Access Token instead. ⚠️ Mange tilbydere (GitHub) støtter ikke lenger passordautentisering. Bruk Access Token i stedet.

⚠️ No embedding providers available. Configure an API key for OpenAI, OpenRouter, or Google in the General tab. ⚠️ Ingen innbyggingsleverandører tilgjengelig. Konfigurer en API-nøkkel for OpenAI, OpenRouter eller Google i fanen Generelt.

⚠️ One-time operation — generates embeddings for all messages. Estimated cost: ~$0.04 for 10,000 messages. ⚠️ Engangsoperasjon — genererer innbygginger for alle meldinger. Estimert kostnad: ~$0,04 for 10 000 meldinger.

⚠️ These are advanced settings ⚠️ Dette er avanserte innstillinger

⚠️ interrupted ⚠️ avbrutt A warning label that appears next to a message that was interrupted.

🌐 🌐 A label for a model that can be used to generate images.

🎨 🎨 A button that generates an image.

👁️ 👁️ A checkmark emoji.

💡 For Gmail, use an App Password. Google Account > Security > 2-Step Verification > App passwords. 💡 For Gmail, bruk et apppassord. Google-konto > Sikkerhet > 2-trinns bekreftelse > Apppassord.

💡 Use HTTPS URL (e.g., https://gitlab.pm/user/repo.git) — works with all auth methods. 💡 Bruk HTTPS URL (f.eks. https://gitlab.pm/user/repo.git) — fungerer med alle autentiseringsmetoder.

🔧 🔧 A checkmark emoji.

oAI oAI Bundle name