Fieldcraft
Fieldcraft is a local-first Chrome extension that reads job pages, evaluates fit, researches the company, and drafts answers you review before applying.
Overview
Fieldcraft is a local-first Chrome extension for job applications. It reads the job description and form fields from the open tab, scores how well the role fits your saved profile, researches the company, and drafts truthful answers that you review before inserting anything.
Fieldcraft is a local-first Chrome extension for job applications. It reads the job description and form fields from the open tab, scores how well the role fits your saved profile, researches the company, and drafts truthful answers that you review before inserting anything.
Everything stays in your browser. The extension talks directly to your chosen AI provider with an API key you control, and it never submits a form for you.
The product is shipped as a Chrome MV3 side-panel extension and a companion Next.js marketing site. The extension uses a Vite + CRXJS build, a React side panel, and background scripts that inject content helpers into job pages.
How It Was Built
The main technical choices behind the product, from system design to the parts that make it work day to day.
- Built a Chrome MV3 side-panel extension with manifest v3, service worker, and content scripts that extract job descriptions and form fields from live job pages.
- Designed a profile-based fit-scoring flow that compares role requirements against user-provided facts, surfaces real gaps, and gathers a concise company brief using web research.
- Built a Chrome MV3 side-panel extension with manifest v3, service worker, and content scripts that extract job descriptions and form fields from live job pages.
- Designed a profile-based fit-scoring flow that compares role requirements against user-provided facts, surfaces real gaps, and gathers a concise company brief using web research.
- Implemented a local-first answer drafting loop where the model generates per-field responses and the user chooses what to insert after explicit review.
- Shipped the public landing page with GSAP animations, semantic HTML, Open Graph/Twitter metadata, JSON-LD, sitemap, robots.txt, and agent-ready discovery files (llms.txt, ai.txt, AGENTS.md).
Impact
- Cuts per-application time by ~55–65% by automating job-description reading, form-field parsing, role-fit scoring, and profile-grounded answer drafting.
- Reduces research and tab context switching by ~70–80% by surfacing a concise company brief, role-fit score, and per-field drafts inside the same side panel.
- Cuts per-application time by ~55–65% by automating job-description reading, form-field parsing, role-fit scoring, and profile-grounded answer drafting.
- Reduces research and tab context switching by ~70–80% by surfacing a concise company brief, role-fit score, and per-field drafts inside the same side panel.
- Keeps all profile data and API keys in the browser with no server-side storage.
- Requires explicit user approval before any drafted answer is inserted, so nothing is submitted without review.
Highlights
- Works inside the job page as a Chrome side panel, not a separate dashboard.
- Drafted answers are reviewed before insertion, keeping the user in control.
- Works inside the job page as a Chrome side panel, not a separate dashboard.
- Drafted answers are reviewed before insertion, keeping the user in control.
- Profile and API keys stay in the browser with no server-side storage.
Tech Stack
More Projects
Additional work across AI products, developer tooling, and full-stack systems.
Inkdown
Markdown editor and sharing platform with folders, live preview, and public links.
DeployNinja
GitHub-native deployment platform for automated builds, live logs, and repeatable releases.
Intellect AI
AI content platform combining image tools, video tools, and community features.