-
Notifications
You must be signed in to change notification settings - Fork 2
Core Features
The landing page offers a resume, an academic CV, and a cover letter. Creating one prepares a fresh document state and opens the matching editor.
A new document also starts a fresh Agent Mode session. Earlier chat and review state do not leak into the new workspace.
The resume editor covers personal details, a professional summary, education, experience, skills, projects, and awards.
English and Chinese document modes are available. The document language is separate from the interface language.
The academic CV editor supports the broader structure expected in scholarly work. It includes research interests, education, research and teaching experience, industry experience, publications, presentations, grants, service, technical skills, and references.
English and Chinese document modes are available here as well.
The cover letter editor focuses on sender details, recipient details, the letter date, and body paragraphs. It currently uses English document mode.
Each editor pairs a structured form with a paginated preview. Form changes appear immediately, so the preview stays useful while writing and not only at export time.
Agent edits use the same document state. Review highlights can show what changed without creating a second copy of the document.
CVForge can import its own JSON document files. It can export the current document as JSON, PNG, or PDF.
PNG and PDF export happen inside the browser. The preview is captured locally and the document is not sent to an export service.
Every editor can load example content. Examples show the expected structure, date style, and writing density without locking users into one template.
The current tab stores document and Agent Mode state in sessionStorage. Refreshing the page keeps the work. Closing the tab starts a clean session next time.
LLM configuration and interface language use localStorage because they are useful across visits.
Agent Mode can answer questions, improve wording, add supported content, review changes, and use uploaded PDF, Markdown, or text references.
Responses stream into the chat panel. Compatible models can also show optional thinking controls and collapsed reasoning content.
CVForge Wiki