Saving content

Everything in Pocketmind starts with a save. Captured content is processed into text, summaries, tags, and embeddings so you can search and chat over it later.

What you can save

• Web articles and blog posts (via URL)
• PDFs and documents (upload + OCR)
• YouTube / Vimeo videos (transcription when supported)
• Audio files you upload, such as podcast episodes (transcription + searchable text; audio links are not fetched yet)
• Images (analysis when enabled)
• Notes you write yourself

Capture methods

Web app

Use capture in the dashboard: paste a URL or upload a file. Processing runs in the background — refresh the library when the item becomes ready.

REST ingest API

Automations can push content (markdown or canonical JSON payloads) with an API key that includes the ingest:write scope. This is not a “fetch this URL” endpoint — your client supplies the payload.

1. In the app, open Settings → API Keys and create a key with ingest:write.
2. Send POST /v1/ingest to your Pocketmind API gateway with Authorization: Bearer <api-key> (or X-API-Key) and a JSON body shaped like:

{
  "requests": [
    {
      "source_id": "newsletter-2026-10-11",
      "format": "markdown",
      "title": "Weekly notes",
      "payload": "# Notes\n\nYour markdown body…"
    }
  ]
}

source_id is your idempotency key (same id upserts). format is markdown or canonical. Confirm the item appears in the library after processing completes.

Mobile PWA

Install Pocketmind as a PWA, then share links into the app from the system share sheet when available. Local PDF/image/audio files are not ingested via share today — upload those from the web app instead.

Chrome extension

One-click save from the browser is marked Soon on the marketing site — use URL capture until it ships.

Agents (MCP)

Connected agents can retrieve (and, depending on permissions, help capture into) your library. Prefer OAuth2; API keys are the fallback. See MCP for agents.

After you save

1. Content is fetched / extracted.
2. AI produces a summary and tags.
3. Embeddings enable semantic search and RAG chat.
4. Items can land in smart collections based on filters you set.

If an item stays stuck on processing, wait a minute and reload. Persistent failures usually mean the source blocked fetching or the file type is unsupported — try a PDF upload or a different URL.