Purl

io.github.nublsonPublished by GitHub account nublsonv0.5.1Updated Nov 7, 2026

Save, organize and share your Purl links from any MCP client.

Overview

AI-generated overview

Purl is a hosted bookmark service whose MCP server lets an assistant save, list, read, move and delete saved links and organize them into folders.

What you can do
  • Save a URL as a link, with automatic detection of web pages, PDFs, YouTube videos and audio and resolved title, description, favicon and thumbnail.
  • List saved items and fetch a single link's details.
  • Move a link between folders.
  • List, create, update and delete folders.
  • Re-saving an existing URL refreshes its metadata and moves it to the top of the list.
When to use it
Useful when you want an assistant to keep a personal collection of links worth revisiting, or to file and tidy that collection into folders while chatting, without opening the Purl app.
Requirements
Remote MCP endpoint at purl.live, reachable over the network. Authentication is required: OAuth, or an Authorization header carrying a Purl API key created in Settings, Integrations. A Purl account is needed, and each account can save up to 1,000 links.
Before you install
The Authorization header carries a secret API key; keep it out of logs and shared transcripts. The tools write to your account: saving, moving, editing and deleting links and folders change stored data, and folder deletion may remove its contents. Saved URLs are fetched and scraped server-side by Purl, so links you add are sent to that service.

Installation

In SourceWeft

  1. Open Purl in the dashboard and add it to a workspace.
  2. Connect the required credentials. They are stored privately and never shown on public pages.
  3. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/io.github.nublson%2Fpurl/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "purl": {
      "type": "http",
      "url": "https://purl.live/api/mcp"
    }
  }
}

README

[Purl — Save Anything. Keep what matters. A home for your links, PDFs, video, and audio.]

Purl

Save the links worth keeping. Share the ones worth passing on.

Live preview: https://purl.live

Purl is a calm place to save and share the links you find — a home for your "pearls". You paste URLs: web pages, PDFs, YouTube videos, and audio. Purl resolves each item's metadata (title, favicon, description, thumbnail) and keeps everything in one place, available from the app, a REST API, and an MCP server.

Purl is free, with one limit: each account can save up to 1,000 links (MAX_SAVED_LINKS in src/lib/limits.ts).

The product goal: one place to keep the links you care about, and an easy way to share the ones worth passing on.

Implemented today

  • Marketing site — Landing page (hero with a live preview), docs for the REST API and MCP server, privacy and terms.
  • Authentication — Google and GitHub sign-in on the landing page via Better Auth (Apple too when its env vars are set). Every account gets a unique username, editable in Settings, where you can also connect or disconnect sign-in methods; avatars come from the OAuth provider's profile photo.
  • Save & organize
    • Add items by URL with automatic content-type detection (web, PDF, YouTube, audio).
    • Links grouped by relative time (e.g. Today, This week, Last month).
    • Preview metadata (title, description, favicon, thumbnail where available).
  • Hardened outbound fetch — Server-side safeFetch with optional proxy/DNS controls (see AGENTS.md). An egress proxy can be configured via SAFE_OUTBOUND_HTTP_PROXY.
  • Realtime list sync — Supabase Realtime so saves and updates propagate across tabs/devices quickly.
  • Link actions — Open original, copy URL, edit metadata, delete.
  • REST API & MCP — /api/v1 (links and folders) and an MCP server (save_link, list_saved_items, get_link, move_link, list_folders, create_folder, update_folder, delete_folder) with API-key or OAuth auth.
  • Operational extras — Optional Upstash-backed API rate limiting, Vitest coverage for critical paths.
  • PWA (installable app) — Web App Manifest plus a Serwist service worker (src/app/sw.ts) that builds to public/sw.js (generated on pnpm build, gitignored). Enables Install in Chrome/Edge and similar where the platform supports it, with runtime caching via Serwist's Next.js defaults and a static offline shell at /~offline. Serwist is disabled in pnpm dev to avoid service-worker cache surprises during development — use pnpm build && pnpm start (or your production URL) to exercise installability and the SW.

Save flow

Saving a link is fully synchronous — there is no background processing.

  1. Input — POST /api/links with a URL (also /api/v1/links and the MCP save_link tool).
  2. Classify & decorate — Server-side detectContentType (SSRF-safe HEAD / sniff) plus scrapeLinkMetadata (Open Graph HTML, PDF Content-Disposition / size, YouTube oEmbed). Saving an existing URL again refreshes its metadata and moves it to the top.
  3. Persist — A Link row with title, favicon, thumbnail, domain, and contentType (WEB, PDF, YOUTUBE, or AUDIO).
  4. Sync — broadcastLinksChanged notifies other tabs/devices via Supabase Realtime.

Not implemented yet

These are called out explicitly because the repo is going public:

  • Settings breadth — Settings cover username, sign-in methods, and account deletion; broader account preferences (notification settings, etc.) are not implemented yet.

Marketing vs. product: The landing page copy mentions ideas such as collections and a weekly digest. Those are not built in the current schema or app — treat them as roadmap, not shipped features.

Tech stack

  • Web: Next.js (App Router), React, TypeScript
  • UI: Tailwind CSS, shadcn/ui
  • Auth: Better Auth
  • Database: PostgreSQL + Prisma
  • Email (optional in dev): Resend for feedback emails
  • Realtime: Supabase client (anon + service role on server)
  • PWA: Serwist (@serwist/next), web manifest + precache / offline fallback

CI / GitHub Actions

Automation lives under .github/workflows/. Every PR and manual release is gated by these pipelines.

PR checks — pr-checks.yml

Runs on pull_request to develop and main: Setup & validation → Prisma (generate client + type fixes) → Lint and type check (in parallel) → Tests and production build (in parallel, after lint and type check pass). Concurrency is per-PR so new pushes cancel stale runs.

[GitHub Actions graph for pr-checks.yml: Setup, Prisma, Lint & Type Check, Test & Build]

Release — release.yml

Runs on workflow_dispatch (manual): Merge develop into main, then build validation so production is only promoted after a green build.

[GitHub Actions graph for release.yml: merge develop into main, then build validation]

Security

Purl is built around untrusted input (arbitrary URLs). A few layers matter in production:

  • SSRF-aware outbound fetches — User-supplied URLs are not passed to raw fetch. OG/thumbnail probes, PDF fetch, content-type sniffing, and similar paths go through safeFetch: HTTP(S) only, blocked private/link-local/reserved targets, redirect handling with per-hop host checks, DNS resolution pinned before connect (mitigates classic DNS rebinding against the pre-check), optional response size caps (e.g. PDF proxy). Optional egress proxy and custom DNS servers are documented in AGENTS.md.
  • Authentication & route gating — Better Auth sessions; Next.js proxy redirects unauthenticated users away from private routes. Sign-in is OAuth-only (Google/GitHub, plus Apple when configured); there are no passwords.
  • API authorization — Sensitive routes (/api/links, /api/v1/*, MCP, etc.) resolve the session server-side and scope work to the signed-in user.
  • Rate limiting — When UPSTASH_REDIS_REST_URL and UPSTASH_REDIS_REST_TOKEN are set, the proxy applies per-IP limits to /api/auth/*, POST /api/links, and POST /api/feedback (see proxy-rate-limit.ts). Without Upstash, limits are disabled — fine locally, not ideal for production.
  • Secrets & client exposure — SUPABASE_SECRET_KEY (or the legacy SUPABASE_SERVICE_ROLE_KEY) and similar values are server-only. The browser uses the Supabase anon key for Realtime only; .env stays gitignored.
  • Response bounds — PDF proxy streaming is size-capped (see safe-outbound-fetch).

Reporting a vulnerability: use GitHub Security Advisories for this repository so details stay private until patched.

Setup (local development)

Prerequisites

  • Node.js: recent LTS
  • Package manager: pnpm (this repo includes pnpm-lock.yaml)
  • Postgres: local or hosted (Supabase works well)

1) Install dependencies

bash
pnpm install

2) Configure environment variables

Create a .env file in the repo root. See .env.example for the full list; minimum for core behavior:

bash
DATABASE_URL="postgresql://USER:PASSWORD@HOST:5432/DBNAME"
# Supabase Realtime — cross-device instant link list sync (same project as Postgres)SUPABASE_URL="https://YOUR_PROJECT.supabase.co"SUPABASE_ANON_KEY="eyJ..."SUPABASE_SECRET_KEY="sb_secret_..."
# Sign-in (OAuth apps with callback {BETTER_AUTH_URL}/api/auth/callback/{google,github})GOOGLE_CLIENT_ID="..."GOOGLE_CLIENT_SECRET="..."GITHUB_CLIENT_ID="..."GITHUB_CLIENT_SECRET="..."
# Optional (used to send in-app feedback emails)RESEND_API_KEY="re_..."RESEND_FROM="Purl <[email protected]>"

Notes:

  • DATABASE_URL is required (Prisma + Better Auth).
  • Supabase env vars are required for realtime link list sync. Use Project Settings → API in the Supabase dashboard. The service role key must stay server-only.
  • Google/GitHub OAuth credentials are required in production; locally, a provider whose vars are missing is simply hidden. For local dev, use callback URLs http://localhost:3000/api/auth/callback/{google,github}. Apple is optional and enabled only when all four APPLE_* vars are set (see .env.example).
  • Resend is optional for local dev: without RESEND_API_KEY, in-app feedback can't be sent.
  • Better Auth secrets and URLs are in .env.example — copy those keys for a working auth setup.

3) Run database migrations

bash
pnpm prisma migrate dev

4) Generate Prisma client (if needed)

bash
pnpm prisma generate

5) Start the dev server

bash
pnpm dev

Open http://localhost:3000.

Optional: a named HTTPS URL with portless. With portless installed globally (npm install -g portless, Node.js 24+), run portless run instead of pnpm dev. It runs the dev script behind a local proxy at https://purl.localhost. Point auth at that URL in .env.local, or sign-in rejects the origin:

bash
BETTER_AUTH_URL="https://purl.localhost"BASE_URL="https://purl.localhost"

In a git worktree, portless prefixes the branch name (for example https://my-branch.purl.localhost); set both variables to that URL there.

PWA / install: With pnpm dev, the service worker is not active. After a production build, public/sw.js exists locally; run pnpm start and open the app in Chromium to use Install or to test offline navigation to /~offline.

Testing

Tests use Vitest and focus on critical logic (formatters, link grouping, auth routing, API behavior, metadata scraping). They intentionally avoid shallow UI-only wrappers.

bash
pnpm test        # run oncepnpm test:watch  # watch mode

Useful commands

bash
pnpm lintpnpm buildpnpm startpnpm test

More contributor notes (Prisma, outbound proxy env): see AGENTS.md.

Source: README.md at commit a140440

Tools

0
This server asks you to sign in, so SourceWeft could not read its tools without an account and has not indexed them. What a connected account sees can depend on that account and its permissions.

Version history

16
  1. v0.5.1LatestNov 7, 2026
  2. v0.3.0Oct 29, 2026
  3. v0.2.0Oct 21, 2026
  4. v0.1.16Oct 15, 2026
  5. v0.1.15Oct 11, 2026
  6. v0.1.14Oct 10, 2026
  7. v0.1.13Oct 9, 2026
  8. v0.1.12Oct 8, 2026
  9. v0.1.11Oct 4, 2026
  10. v0.1.10Oct 3, 2026
  11. v0.1.9Oct 2, 2026
  12. v0.1.8Oct 2, 2026
  13. v0.1.6Oct 1, 2026
  14. v0.1.4Sep 25, 2026
  15. v0.1.3Sep 24, 2026
  16. v0.1.1Sep 16, 2026

Score

75% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

    Declare a license in the repository or the registry entry.