Turn any company URL into a structured competitive-intelligence dossier — without leaving VS Code.
Give Company Dossier a company name and URL. It crawls the live site, reconstructs its history from the Wayback Machine, fingerprints DNS/email infrastructure and the tech stack, pulls public records, and — optionally — uses Claude to synthesize an executive brief, SWOT, and ranked findings. The result is a navigable, agent-friendly folder of Markdown written right into your workspace. Public sources only.
- One-input research — give a company name + URL, get a complete dossier.
- Live website crawl — homepage, sitemap, and up to 50 prioritized pages (about, team, products, pricing, contact, news).
- Wayback history — multi-query CDX recon: total captures, unique URLs, archived PDFs, and deleted pages.
- DNS & email recon — MX provider, SPF, DMARC, subdomains, verification tokens.
- Tech-stack detection — CMS, frameworks, CDN, analytics IDs, GTM, and ad pixels.
- Public records — USASpending federal contracts and discovered social profiles.
- AI synthesis (optional) — executive brief, SWOT, ranked findings, and "manual research needed", grounded in collected evidence (requires an Anthropic API key).
- Structured output — a 12-section dossier with YAML frontmatter,
_MOC.mdmaps of content, and aROUTER.mdso AI agents can navigate in two reads. - Public data only — no logins, no scraping behind authentication.
From the Marketplace
- Open VS Code → Extensions (
Ctrl/Cmd+Shift+X). - Search for Company Dossier — Intelligence Research.
- Click Install, or open it on the Marketplace.
From source
git clone https://github.com/ever-just/company-dossier-vscode
cd company-dossier-vscode
npm install
npm run compileThen press F5 in VS Code to launch an Extension Development Host, or package a .vsix
with npx vsce package and install it via Extensions → "Install from VSIX…".
Open a folder in VS Code first — dossiers are written into the open workspace.
- Sidebar view — click the Company Dossier icon in the Activity Bar, enter a company name + URL, pick a depth, and run the pipeline with live progress.
- Chat participant — in the Chat view (Copilot Chat), type:
Results stream back inline with a button to open the generated dossier.
@dossier /research Acme Corp https://acme.com - Command Palette (
Ctrl/Cmd+Shift+P):- Dossier: Research Company — prompts for name + URL, then runs the full pipeline.
- Dossier: New Entity Profile — scaffolds a profile (person, supplier, competitor, client, product) into the right section.
Optional AI synthesis: add your key in Settings → Company Dossier → Anthropic API Key
(companyDossier.anthropicApiKey). Without a key, data is still collected and written; only
the LLM synthesis step is skipped. Tune crawl depth with companyDossier.maxPages.
Each run creates a "<Company> DOSSIER/" folder in your workspace:
<Company> DOSSIER/
├── README.md ← Landing page
├── ROUTER.md ← AI-agent navigation (question → file)
├── CHANGELOG.md
├── _meta/ ← Methodology, confidence scale
├── _data/ ← Structured data (CSVs)
├── _assets/ ← Photos & PDFs
├── _evidence/ ← Raw source data
├── 1_corporate/ ← Identity, legal, tech stack, contacts
├── 2_people/profiles/ ← One file per person
├── 3_products/ ← Products, services, pricing
├── 4_suppliers/profiles/ ← One file per supplier
├── 5_customers/ ← Clients, prospects
├── 6_competitors/profiles/ ← One file per competitor
├── 7_financials/ ← Revenue, valuation, federal contracts
├── 8_marketing/ ← Social, events, press
├── 9_brand/ ← Visual identity
├── 10_timeline/ ← History, website evolution
├── 11_analysis/ ← Executive brief, SWOT, key findings
└── 12_industry/ ← Market context
Every file carries YAML frontmatter (type, confidence, last-updated). With an API key, the analysis section is populated with an executive brief, SWOT, key findings, and a manual-research checklist.
Company Dossier is one part of a larger ecosystem:
- Website — companydossier.lol
- npm package —
company-dossier— run the same research from your terminal:npx company-dossier <company>
- Core repo & methodology — github.com/ever-just/company-dossier
- Architecture — the collector → generator flow and pipeline phases.
- AGENTS.md — how an AI agent should understand and extend this repo.
- Contributing · Code of Conduct · Security · Changelog
MIT © EverJust