Hi Mohamed,

This is indeed a very interesting app!

I am suggesting a feature: disable new users sign-up

Usecase

After the app is deployed and exposed to internet, it is possible to sign-up. However, it would be nice to prevent other people from signing up to one's instance.

Thanks for your work.

The mobile app is already cross-platform. We just need to test and publish it to the play store.

When “Add to home page” is selected on IOS from mobile safari, the icon is solid black.

This is really great software, I love the clear motivation and where it fits into the ecosystem.

I'd be lovely to get an API for some common actions, primarily retrieving content, searching, and addition from outside the API.

Use case:

• Allow interaction from external channels such as slack/discord bots
• Ability to integrate with AI orchestration tooling (like mine - https://github.com/omnitool-ai/omnitool)

Hi,

Could you add support for HTTP and SOCKS5 proxies?

Many users, including myself, need proxies to access sites like Google and Facebook due to local internet restrictions.
This feature would help a lot.

Thanks,

I have this error inside the Hoarder-WORKERS container:
openai job failed: MeiliSearchApiError: The Authorization header is missing. It must use the bearer authorization method.

And i get this error inside the Hoarder-MEILI container:

MeiliSearchApiError: The Authorization header is missing. It must use the bearer authorization method.
    at /app/apps/web/.next/server/chunks/673.js:4817:1188
    ... 2 lines matching cause stack trace ...
    at process.processTicksAndRejections (node:internal/process/task_queues:95:5) {
  code: INTERNAL_SERVER_ERROR',
  name: 'TRPCError',
  [cause]: l [MeiliSearchApiError]: The Authorization header is missing. It must use the bearer authorization method.
      at /app/apps/web/.next/server/chunks/673.js:4817:1188
      at Generator.next (<anonymous>)
      at s (/app/apps/web/.next/server/chunks/673.js:4815:69399)
      at process.processTicksAndRejections (node:internal/process/task_queues:95:5) {
    code: 'missing_authorization_header',
    type: 'auth',
    link: 'https://docs.meilisearch.com/errors#missing_authorization_header',
    httpStatus: 401
  }
}

MacOS doesn't like empty named .env files, so when running the docker script I needed to give a full name and modify the compose file to reflect a hoarder.env file

This issue to track supporting ollama as an alternative to openai.

NOTE: not a high priority issue, just reporting what I'm seeing.

Upon adding a new bookmark it failed to crawl the site with the message: {"message":"INVALID_URL, Need to provide a valid URL.","code":"INVALID_URL","description":"Need to provide a valid URL."}

The url was https://www.healthline.com/health/fitness-exercise/plantar-fasciitis-stretches#What-is-plantar-fasciitis?

The mellisearch is also a questionable choice.

30$a month just for search on a self hosted app?

With obsidian this isn't required.

Really like your tool.

I was able to add ollama for AI work, but I'm a bit lost for Android.

How to use the bookmarking tool/app in android?

Thank you.

First off - looks amazing and I'm looking forward trying to replace a couple of my apps with this. (Wallabag, Google Keep, Potentially Plainpad, and some TidddlyWiki use cases)

If I could request one feature then it would be a checklist feature.

Haven't tested the Android app, but would be neat to have also the app in the share dialog to save links/notes.

Good luck with the project, it already looks great! :)

I set the .env variable to DISABLE_SIGNUPS=true but it didn't disable the option.

I ran docker compose up -s again after saving the .env already.

Great repo BTW! Thank you!

Once I've added my bookmark, I'd like to manually edit the title of the item. Yes, the scraper does grab the title text from the metadata but I'd like to change it myself.

Is this possible?

Also, is it possible to import my list of bookmarks from Chrome or another browser?

Each time the browser extension is clicked, a new bookmark is created. If the page has been previously bookmarked, a new bookmark shouldn't be created, or user should be prompted (if there is a need for duplicates - not sure why that may be useful though?)

If I paste a bookmark into the Upper Left box the button on the bottom changes to "Press ⌘+Enter to Save". This is on a Windows machine. I do not have a command key, and no combination of Ctrl or Alt seems to work. I can still press the button with the mouse.

I added ollama support per instructions. Some of the requests seemed to work fine but I found the following in the logs:
2024-03-27T14:01:40.649Z error: [inference][17] inference job failed: Error: Expected a completed response. 2024-03-27T14:06:40.997Z error: [inference][18] inference job failed: TypeError: fetch failed

At least one of my bookmarks does not have any tags added to it.

I'm running this on a docker container on linux.

when adding bookmarks from the chrome extension, it would be nice for the app to auto-update to changes so you dont need to refresh

ummm... there needs to be an Undelete link here!! I.e., The bookmark you didn't mean to delete has been deleted! Undelete?
⋮
I looked at the screen shot again: The bookmark has been deleted! Undelete? would be good too.

Couple other things... and they may exist, don't know yet, but audiobookshelf has a backup system where you can keep n number of days of backups of the database. If you don't have that, then that would be highly desirable.

Is there a log of the url I deleted? I have no idea what it was. I hope it wasn't the one about cleaning mortar off bricks in new zeland!

Erroring out on a nonstandard port.

Redis supports 16 different databases, numbered from 0 to 15. By default, clients connect to database 0.

If the database number was configurable I could use my already running Redis instance, instead of spinning up a new one.

It seems like this project uses BullMQ, which itself uses ioredis. The configuration option db is the "Database index to use.".

More information: https://github.com/redis/ioredis/blob/v4/API.md#new-redisport-host-options

the readme suggests this functionality exists but it doesnt seem to...

Testing out this app on my homelab. Testing uploads of various images and i ran into an issue with a 17MB PNG File.

Hello,
I set up Hoarder behind cosmos using the latest version, and the service is working fine using OpenRouter too for free inference.
But, when trying to use the chrome extension I get a "Failed to fetch" error. Looking at the cosmos logs, I see:

2024/03/30 15:48:44 "OPTIONS https://***.xyz/api/trpc/apiKeys HTTP/2.0" from 192.168.1.1:40518 - 204 0B in 3.163343ms

when I try to xh the same link I get:

HTTP/2.0 404 Not Found
access-control-allow-credentials: true
access-control-allow-credentials: true
access-control-allow-headers: Content-Type, Authorization
access-control-allow-methods: GET, POST, PUT, DELETE, OPTIONS
access-control-allow-origin: *.xyz
access-control-allow-origin: *
content-security-policy: frame-ancestors 'self'
content-type: application/json
date: Sat, 30 Mar 2024 15:59:28 GMT
strict-transport-security: max-age=31536000; includeSubDomains
vary: RSC, Next-Router-State-Tree, Next-Router-Prefetch, Next-Url
x-content-type-options: nosniff
x-ratelimit-limit: 6000
x-ratelimit-remaining: 86
x-ratelimit-reset: 1711815568
x-served-by-cosmos: 1
x-timeout-duration: 4h0m0s
x-xss-protection: 1; mode=block

{
    "error": {
        "json": {
            "message": "No \"query\"-procedure on path \"apiKeys\"",
            "code": -32004,
            "data": {
                "code": "NOT_FOUND",
                "httpStatus": 404,
                "path": "apiKeys",
                "zodError": null
            }
        }
    }
}

and in the hoarder-web logs I have:

[next-auth][warn][NEXTAUTH_URL] 
https://next-auth.js.org/warnings#nextauth_url
s [TRPCError]: No "query"-procedure on path "apiKeys"
    at m (/app/apps/web/.next/server/chunks/673.js:4826:4202)
    at /app/apps/web/.next/server/app/api/trpc/[trpc]/route.js:1:4251
    at Array.map (<anonymous>)
    at g (/app/apps/web/.next/server/app/api/trpc/[trpc]/route.js:1:4185)
    at process.processTicksAndRejections (node:internal/process/task_queues:95:5) {
  code: 'NOT_FOUND',
  [cause]: undefined
}

The full docker compose for the web app (hiding envs) is:

{
  "services": {
    "Hoarder-WEB": {
      "container_name": "Hoarder-WEB",
      "image": "ghcr.io/mohamedbassem/hoarder-web:latest",
      "environment": [
        ***
      ],
      "labels": {
        "cosmos-auto-update": "true",
        "cosmos-force-network-mode": "cosmos-web-default",
        "cosmos.stack": "web",
        "cosmos.stack.main": "true"
      },
      "ports": [
        "0.0.0.0:8096:3000/tcp",
        ":::8096:3000/tcp"
      ],
      "volumes": [
        {
          "Type": "bind",
          "Source": "/volume1/docker/hoarder/data",
          "Target": "/data"
        }
      ],
      "networks": {
        "cosmos-web-default": {},
        "hoarder": {}
      },
      "routes": null,
      "restart": "on-failure",
      "devices": null,
      "expose": [],
      "depends_on": [],
      "command": "/bin/sh -c (cd /db_migrations && node index.js) && node server.js",
      "entrypoint": "docker-entrypoint.sh",
      "working_dir": "/app/apps/web",
      "user": "root",
      "hostname": "26308f0b7bf2",
      "network_mode": "cosmos-web-default",
      "healthcheck": {
        "test": null,
        "interval": 0,
        "timeout": 0,
        "retries": 0,
        "start_period": 0
      }
    }
  }
}

Minor display bug but Favourites (or Favorites in US English) is spelled as Favourties in the left Nav menu.

Attempted to upload a webp image, it fails due to unsupported file type.

When search is selected or characters are entered in the search box the padding on the sides of the column are reduced.

It’s actually better in the search mode, so applying that style to the main view and reducing the padding would be appropriate.

I installed Hoarder on my Laptop for test purposes, and now i wonder how i can access it.

Adding a request for a Safari plugin

This is what it currently looks like when bookmarking a page:

Would be cool to let it run on a Raspberry. Thanks.

Just a couple of ideas to remove the cloud of tags that appears with OpenAI tagging.

Would it be possible to have OpenAI look at existing tags and and choose from the most relevant ones? Perhaps this could be done more efficiently with an embeddings search on the documents and/or on the existing tags to find similar ones.

Another Idea could be to set a "number of tags to generate" with OpenAI so it could settle for an existing general one vs. something new and hyper specific.

Eg. This link generated some pretty silly tags in my opinion

When you delete an entry on /dashboard/search it shows The bookmark has been deleted! but doesn't remove it from the list like it does on /dashboard/bookmarks.

When I try to connect to my instance of hoarder on a Tailscale IP and a non standard port, it works on a webpage but doesn't seem to work in a browser extension:

Web page:

Browser extension:

Is it possible to handle websites with an overlay for cookies better?
Many websites i add is shows no thumbnail and shows as the text only the cookies informations.

I'm attempting to add a text bookmark (markdown specifically) and I keep getting a red warning at the bottom of the screen saying "something went wrong". Which docker log should I look in for additional logging messages?

My text if it helps, just my own notes from doing cloudflare dns.

Cloudflared DNS

https://docs.pi-hole.net/guides/dns/cloudflared/

https://serverfault.com/questions/956447/bind-set-port-for-forwarders

AMD64 architecture (most devices)¶
Download the installer package, then use apt-get to install the package along with any dependencies. Proceed to run the binary with the -v flag to check it is all working:

For Debian/Ubuntu

wget https://github.com/cloudflare/cloudflared/releases/latest/download/cloudflared-linux-amd64.deb
sudo apt-get install ./cloudflared-linux-amd64.deb
cloudflared -v

Configuring cloudflared to run on startup¶
Create a cloudflared user to run the daemon:

sudo useradd -s /usr/sbin/nologin -r -M cloudflared
Proceed to create a configuration file for cloudflared:

sudo nano /etc/default/cloudflared
Edit configuration file by copying the following in to /etc/default/cloudflared. This file contains the command-line options that get passed to cloudflared on startup:

# Commandline args for cloudflared, using Cloudflare DNS
CLOUDFLARED_OPTS=--port 5053 --upstream https://1.1.1.1/dns-query --upstream https://1.0.0.1/dns-query

Update the permissions for the configuration file and cloudflared binary to allow access for the cloudflared user:

sudo chown cloudflared:cloudflared /etc/default/cloudflared
sudo chown cloudflared:cloudflared /usr/local/bin/cloudflared

Then create the systemd script by copying the following into /etc/systemd/system/cloudflared.service. This will control the running of the service and allow it to run on startup:

sudo nano /etc/systemd/system/cloudflared.service

[Unit]
Description=cloudflared DNS over HTTPS proxy
After=syslog.target network-online.target

[Service]
Type=simple
User=cloudflared
EnvironmentFile=/etc/default/cloudflared
ExecStart=/usr/local/bin/cloudflared proxy-dns $CLOUDFLARED_OPTS
Restart=on-failure
RestartSec=10
KillMode=process

[Install]
WantedBy=multi-user.target

Enable the systemd service to run on startup, then start the service and check its status:

sudo systemctl enable cloudflared
sudo systemctl start cloudflared
sudo systemctl status cloudflared

Now test that it is working! Run the following dig command, a response should be returned similar to the one below:

pi@raspberrypi:~ $ dig @127.0.0.1 -p 5053 google.com

; <<>> DiG 9.11.5-P4-5.1-Raspbian <<>> @127.0.0.1 -p 5053 google.com
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 12157
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
; COOKIE: 22179adb227cd67b (echoed)
;; QUESTION SECTION:
;google.com.                    IN      A

;; ANSWER SECTION:
google.com.             191     IN      A       172.217.22.14

;; Query time: 0 msec
;; SERVER: 127.0.0.1#5053(127.0.0.1)
;; WHEN: Wed Dec 04 09:29:50 EET 2019
;; MSG SIZE  rcvd: 77

when i paste or drag an image into the "save" box, nothing happens. no console errors. chrome on mac

EDIT: i see it works from mac screenshot thing, but not from copy/paste or dragging from chrome. how is this feature intended to be used?

Hi just got this setup with Ollama for inference. I'm seeing the above error in the workers container when I click the "Recrawl All Links" button on the admin page. Also I get this series of errors adding a note:

2024-03-30T18:38:18.740Z error: [inference][3] inference job failed: SqliteError: no such table: bookmarks
2024-03-30T18:38:18.742Z error: Something went wrong when marking the tagging status: SqliteError: no such table: bookmarks
2024-03-30T18:38:18.760Z error: [search][5] openai job failed: SqliteError: no such table: bookmarks
2024-03-30T18:38:19.245Z error: [inference][3] inference job failed: SqliteError: no such table: bookmarks
2024-03-30T18:38:19.246Z error: Something went wrong when marking the tagging status: SqliteError: no such table: bookmarks
2024-03-30T18:38:19.851Z error: [search][5] openai job failed: SqliteError: no such table: bookmarks
2024-03-30T18:38:20.250Z error: [inference][3] inference job failed: SqliteError: no such table: bookmarks
2024-03-30T18:38:20.250Z error: Something went wrong when marking the tagging status: SqliteError: no such table: bookmarks
2024-03-30T18:38:21.956Z error: [search][5] openai job failed: SqliteError: no such table: bookmarks
2024-03-30T18:38:25.971Z error: [search][5] openai job failed: SqliteError: no such table: bookmarks
2024-03-30T18:38:33.996Z error: [search][5] openai job failed: SqliteError: no such table: bookmarks

Here is my docker-compose file:

version: "3.8"
services:
  web:
    image: ghcr.io/mohamedbassem/hoarder-web:${HOARDER_VERSION:-release}
    restart: unless-stopped
    volumes:
      - /docker-data/hoarder/data:/data
    ports:
      - 81:3000
    env_file:
      - stackstack.env
    environment:
      REDIS_HOST: redis
      MEILI_ADDR: http://meilisearch:7700
      DATA_DIR: /data
  redis:
    image: redis:7.2-alpine
    restart: unless-stopped
    volumes:
      - /docker-data/hoarder/redis:/data
  chrome:
    image: gcr.io/zenika-hub/alpine-chrome:100
    restart: unless-stopped
    command:
      - --no-sandbox
      - --disable-gpu
      - --remote-debugging-address=0.0.0.0
      - --remote-debugging-port=9222
  meilisearch:
    image: getmeili/meilisearch:v1.6
    restart: unless-stopped
    env_file:
      - stack.env
    volumes:
      - /docker-data/hoarder/meilisearch:/meili_data
  workers:
    image: ghcr.io/mohamedbassem/hoarder-workers:${HOARDER_VERSION:-release}
    restart: unless-stopped
    volumes:
      - /docker-data/hoarder/workers:/data
    env_file:
      - stack.env
    environment:
      REDIS_HOST: redis
      MEILI_ADDR: http://meilisearch:7700
      BROWSER_WEB_URL: http://chrome:9222
      DATA_DIR: /data
      # OPENAI_API_KEY: ...
    depends_on:
      web:
        condition: service_started

volumes:
  redis:
  meilisearch:
  data:

Hi,

I've got Hoarder up and running on my virtual machine running docker. I only access it through IP:PORT.
When signing out, it tries to go back on LOCALHOST:PORT, which won't work.

I've tried adding the API_URL value in the docker-compose.yml file, but no luck.

Other than that: very interesting app, looking forward to using it and seeing it devlop!

Thanks!

Please explain what is supposed to populate in HOARDER_VERSION. You don't mention that here and it's not in the links there.... am I supposed to guess what it is? Should I guess that it supposed to be "release?" Should I guess that I go to the git and grab the current version? If the last, am I to change that every time docker changes the version? This is confusing and thus frustrating...

https://docs.hoarder.app/installation

3. Populate the environment variables

To configure the app, create a .env file in the directory and add this minimal env file:

NEXTAUTH_SECRET=super_random_string
HOARDER_VERSION=release
MEILI_MASTER_KEY=another_random_string

You should change the random strings. You can use openssl rand -base64 36 to generate the random strings.

Persistent storage and the wiring between the different services is already taken care of in the docker compose file.

Keep in mind that every time you change the .env file, you'll need to re-run docker compose up.

If you want more config params, check the config documentation here.

As described it would be awesome if we could share notes with other users on the server, for example share a list with a user and put the notes meant to be shared into said list.

This is a feature missing in usememos/memos and i think it would be great if you could implement it into Hoarder.

When you use a Server Address ending with a slash (https://sub.domain.tld/) the extension shows Failed to fetch.

Dev tools logs:

Access to fetch at 'https://sub.domain.tld//api/trpc/apiKeys.exchange?batch=1' from origin 'chrome-extension://kgcjekpmcjjogibpjebkhaanilehneje' has been blocked by CORS policy: Response to preflight request doesn't pass access control check: Redirect is not allowed for a preflight request.

Removing the slash fixes the issue.

It would be neat if it was possible to define rules that add tags based on metadata.

For example:
https://www.bukowskis.com/sv/lots/1540548-matbord-bento-hem-2000-tal
I'd want the tag furniture added because the URL contains bukowskis.com.

https://www.nordiskagalleriet.se/nemo-lighting/escargot?variant=10213172
I'd want the tag mid-century-modern added because the body contains the text Le Corbusier.

Building on-top that, it would be nice if it was possible to trigger changes in the image tag extraction prompt based on existing tags.

So if the tag furniture is present from the text extraction pass (or through metadata rules) the image prompt would append suggestions focusing on describing the item vs more generic tags.

As the title says, when I logout of Hoarder, it redirects users to localhost:3000. Is there somewhere where this default redirection address can be changed?

I'd like to be able to capture a pdf url, eg https://gavinadair.files.wordpress.com/2017/03/baker-changes-of-mind.pdf

currently, it is captured but no tags or added nor is text extracted

logs:

hoarder-workers 2024-03-27T16:53:27.624Z info: [Crawler][9] Will crawl "https://gavinadair.files.wordpress.com/2017/03/baker-changes-of-mind.pdf" for link with id "h03n4dihn2gp0kn8giwiyir7"                                                                                           hoarder-workers 2024-03-27T16:53:27.813Z info: [search][30] Completed successfully                                                                                                                                                                                                      hoarder-workers 2024-03-27T16:53:27.822Z error: [Crawler][9] Crawling job failed: {}                               

Hey again,

how do i set up auto Tagging in Hoarder?

Thanks