Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/CristianParadaLopez/cv-builder/llms.txt

Use this file to discover all available pages before exploring further.

Skillara AI offers two distinct AI assistance features that work at different stages of the CV creation process. Inline field suggestions help you while you are filling in the form — they transform your informal notes into polished professional text before the CV is even generated. PromptBar editing is available after generation and lets you modify the visual design, layout, or content of the finished CV by typing a plain language instruction.

​
Inline AI field suggestions

The AISuggestField component appears next to supported text fields in the form as a button labelled ✨ ¿No sabés qué poner? La IA te ayuda. Clicking it opens a modal where you describe your experience informally — exactly as you would say it out loud — and the AI rewrites it in professional CV language.

​
Supported contexts

The component accepts a context prop that tells the AI what kind of content it is transforming. The supported context values are:
ContextUsed onWhat the AI produces
experienceWork experience descriptionAction-verb past-tense sentences with metrics where possible
summaryProfessional summary3–4 line professional profile paragraph
educationEducation descriptionAcademic achievements, extracurriculars, and developed skills
skillsSkills section4–6 corporate-language competency names, one per line
toolsTools sectionOfficial tool names (e.g. “Microsoft Excel” not “el excel”), one per line
projectProject descriptionProblem → technologies used → result, max 3 sentences
certificationCertification descriptionWhat skills the certification validates and why it is relevant
volunteerVolunteer work descriptionImpact-focused action sentences for volunteer and social hours

​
How a suggestion is generated

When you click Generar versión profesional, the AISuggestField component sends a POST request to /api/cv/suggest with the following body: 
{
  "userText": "your informal text here",
  "context": "experience",
  "systemPrompt": "Eres un experto en recursos humanos...",
  "examples": [
    "Ejemplo de entrada: \"Ayudé a mi tío en construcción 2 años...\"",
    "Ejemplo de salida: \"Colaboré en proyectos de construcción residencial...\""
  ],
  "contextData": {
    "title": "Desarrollador Full Stack",
    "company": "Acme Corp",
    "position": "Backend Developer"
  }
}
The backend’s suggestField function passes this to the AI model and returns { "suggestion": "..." }. The transformed text is displayed in the modal for you to review. Limits: userText is capped at 600 characters. The character counter is shown beneath the textarea in real time.

​
Example transformation

Here is what the AI does with an informal experience description:
You type: "trabajé en ventas en un supermercado, atendía clientes y cobraba en caja"AI generates: "Gestioné operaciones de punto de venta y atención al cliente en entorno de alto volumen, procesando transacciones y garantizando una experiencia de compra satisfactoria."
If you are not happy with the result, click Otra versión to generate a different phrasing without changing your original input. Each new attempt is labelled with a version number. Once you accept a suggestion by clicking Usar esta versión, the text is written directly into the corresponding form field and you can continue editing it manually.

​
Using existing field content

If you already have some text written in a field when you click ✨, the component pre-populates the modal textarea with (Tengo esto escrito: "...") so the AI can use your existing draft as context rather than starting from scratch.

​
PromptBar: post-generation CV editing

After your CV has been generated and is displayed in the preview, the PromptBar component appears below it. This gives you a natural language interface to modify any aspect of the finished CV — layout, colors, fonts, section order, or content — without going back to the form.

​
How it works

Type an instruction in the input field and press Enter or click Aplicar. The Builder calls handleEdit, which sends a POST request to /api/cv/edit with: 
{
  "currentHTML": "<full CV HTML string>",
  "prompt": "your instruction here"
}
The AI receives the entire current CV HTML and your instruction, modifies only what the instruction specifies, and returns the updated HTML. The Builder replaces the current preview with the new result immediately. Prompt length limit: The backend enforces a maximum of 1,000 characters for the prompt field. Instructions longer than this are rejected with a validation error. Photo handling: If your CV includes a profile photo, the base64 image data is stripped from the HTML before it is sent to the AI (replaced with the string PHOTO_PLACEHOLDER) and re-injected into the response after it returns. This keeps the request payload small and prevents the photo data from interfering with the AI’s HTML editing.

​
Example prompts

Here are some effective instructions you can use with the PromptBar: 
Cambia el color de la barra lateral a verde oscuro

Agrega una sección de idiomas con inglés avanzado y español nativo

Usa una fuente más moderna y aumenta el tamaño de los títulos de sección

Mueve los datos de contacto al encabezado en lugar del sidebar

Hazlo más compacto para que entre en una sola página

Cambia el estilo a minimalista: elimina los colores y la barra lateral

​
Quick suggestion chips

The PromptBar also shows three quick-suggestion buttons beneath the input:
  • Haz los colores más vibrantes
  • Usa una tipografía más moderna
  • Haz el diseño más minimalista
Clicking any of these pre-fills the input field so you can either send it as-is or customize it further before applying.
The PromptBar sends the entire CV HTML to the AI for every edit, and the AI rewrites the complete document each time. If you want to preserve a specific phrase, wording, or data point exactly as it appears, be explicit in your instruction. For example: "Cambia el color del sidebar a azul marino pero no modifiques ningún texto de las secciones".

​
Iterating on edits

Each time you apply a PromptBar instruction, the result becomes the new “current HTML” for any subsequent edits. You can chain multiple instructions — for example, first adjusting the color scheme, then changing the font, then reordering sections — and each edit builds on the previous one. The useCV hook stores the latest HTML in both React state and localStorage so the most recent version is always preserved.

Build docs developers (and LLMs) love

Get started for freeTalk to us