Chrome Extension

Save ChatGPT and Claude conversations, capture public webpages by URL, or select page content and store it as a Context Repo document.

The Context Repo Chrome Extension has three capture paths:

ChatGPT and Claude: Save this chat extracts the visible conversation and creates a text document.
Ordinary HTTP(S) pages: Capture this page sends the URL to Context Repo's Firecrawl-backed scrape endpoint.
Selected page content: Pick element lets you choose part of a supported page and creates a document from the extracted text.

After a successful write, the document is available on the next dashboard, MCP, REST, or ChatGPT App read. Semantic and deep-search availability follows the asynchronous chunking and embedding pipeline.

Installation

Install from the Chrome Web Store

Visit the Context Repo Chrome Extension listing and click Add to Chrome. The extension icon appears in your browser toolbar.

Generate an API key

Go to Dashboard > Settings > API Keys and click New API Key. Give it a name like "Chrome Extension" and pick the Write option for Documents & Collections (the extension needs documents.write to save scraped pages — this scope already includes documents.read).

The full API key is shown once after creation. Copy it immediately — you won't be able to see it again. If you lose it, revoke the key and create a new one.

Configure the extension

Click the Context Repo icon in your browser toolbar to open the extension popup. Paste your API key into the key field and click Connect. The key is stored locally in the browser.

Make your first capture

Open a ChatGPT or Claude conversation and choose Save this chat, or open an ordinary HTTP(S) page and choose Capture this page. Use Pick element when you need text that the URL scraper cannot access, such as content visible only in your signed-in browser session.

How Capture Works

The processing path depends on the capture mode:

Chat capture — The extension's ChatGPT or Claude extractor reads the conversation from the page and converts it to Markdown.
URL capture — Context Repo sends the current public URL through Firecrawl and stores the extracted content.
Element capture — The in-page picker converts the selected element to Markdown and submits it directly.
Indexing — Context Repo schedules hierarchical chunking and embeddings after the document write. Search by title or preview may work before deeper chunks are ready.

The capture response links to the created document. You can then edit it or add it to a collection from the dashboard.

What You Can Capture

URL capture supports ordinary public HTTP(S) pages. Chat capture is specialized for ChatGPT and Claude, while element capture is useful for page content already rendered in the active tab. Common use cases include:

Technical documentation — API references, framework guides, language docs
Articles and blog posts — Tutorials, research, opinion pieces
Stack Overflow answers — Save solutions you've found
GitHub READMEs — Capture project documentation
News and research — Industry reports, product announcements

Once a document is written, it is available on the next read from your dashboard, the 29-tool MCP Server, REST API, or ChatGPT App, subject to the connection's authentication and API-key scopes.

File Upload Formats

Beyond web capture, Context Repo supports uploading files directly through the dashboard in 75+ formats. The Chrome Extension focuses on webpage capture, but your uploaded documents live in the same workspace alongside captured pages.

Supported categories include:

Documents — PDF, DOCX, DOC, RTF, Pages
Presentations — PPTX, PPT, Keynote, ODP
Spreadsheets — XLSX, XLS, CSV, TSV, Numbers
Images — JPG, PNG, GIF, SVG, WebP (with OCR text extraction)
E-Books — EPUB
Web Content — HTML, HTM

See the Documents page for the full upload experience, including manual entry and web scraping with advanced extraction options.

Integration With Your Workflow

Captured pages flow into the same workspace that powers all your Context Repo tools:

MCP Server — Your AI assistant in Claude, Cursor, VS Code, or another connected MCP client can search and read captured content directly.
Search — Find captured content using semantic search (meaning-based) or keyword search (exact matching).
ChatGPT App — Access captured documents from inside ChatGPT conversations.

Troubleshooting

Extension icon doesn't appear

After installing, you may need to pin the extension. Click the puzzle-piece icon in Chrome's toolbar, find Context Repo, and click the pin icon.

"Invalid API key" error

Make sure you copied the full key — it starts with gm_. If you're not sure, go to Dashboard > Settings > API Keys, revoke the old key, and create a new one with Write access on Documents & Collections.

Page doesn't capture

Firecrawl cannot use your browser's signed-in session, and Chrome blocks extensions on internal pages such as chrome://. On an ordinary HTTP(S) page that is visible in your browser but unavailable to Firecrawl, try Pick element. ChatGPT and Claude conversations should use Save this chat.