- Brand
- Beebyte
- Category
- Developer Tools
- Primary Subcategory
- App Hosting & Deployment Platforms
Integration details
Description
Manage your Beebyte account and services. Use ChatGPT to create and manage your websites, inspect hosting settings, manage website files, and work with your databases. Create and deploy static HTML or PHP websites using Beebyte hosting. Access follows the permissions of your authorised Beebyte account.
- Integration type
- Plugin
- Verification status
- Not applicable
- Platform
- ChatGPT
- Primary Subcategory
- App Hosting & Deployment Platforms
- Secondary Subcategories
- None listed
- Brand
- Beebyte
- Access
- Account required
- First tracked
- 2026-11-20
- Tool count
- 28
- Market
- US
The Primary Subcategory used for this profile’s headline score.
Other Subcategories where the Integration is listed.
Not discoverable in ChatGPT yet
ChatGPT now suggests Plugins on its own when they match a user's request, but not every Plugin is being surfaced yet.We haven’t seen Beebyte suggested yet. Sign up and we’ll track it, and your Discoverability Score will appear here once ChatGPT starts surfacing it.
What discovery looks like

Get alerts for Beebyte
Get updates when Beebyte’s Discoverability Score or category rank changes.
Competing in ChatGPT App Hosting & Deployment Platforms
View Category28 tools agents can invoke
Add a DNS record to a domain whose DNS beebyte hosts. name is @ (or empty) for the domain itself, a name relative to the domain (eg. www for www.example.com) or a full name in the domain. data is the record's value, as in list_dns_records: eg. 192.0.2.1 for A, www.example.net for CNAME, "10 mail.example.com" for MX, "10 5 5060 sip.example.com" for SRV (with a name like _sip._tcp), and the text without quotes for TXT (eg. "v=spf1 mx ~all"). ALIAS works like CNAME but can be used for the domain itself, where CNAME can't. A CNAME can't share its name with other records. ttl is in seconds, default the domain's default (usually 1800). Records are public and take effect within a few minutes, but resolvers may cache old values for up to the previous record's TTL. To point a domain or name at a website, add A and AAAA records with the website's ipv4_address and ipv6_address (list_websites); the domain must also be added to the website in the portal. Check list_dns_records first to avoid duplicates, and confirm changes to records that are in use (eg. MX records, which decide where email goes) with the user.
add_dns_record
Copy a file, or a directory with its content, within a website. Missing parent directories of the destination are created. An existing file is only replaced with overwrite; with overwrite a directory is merged into an existing one. Symlinks and special files are skipped (counted in skipped). Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
copy_website_file
Order a new website in an organisation (ids are returned by list_organisations). The website is a paid service, billed monthly to the organisation. Confirm with the user before creating one, and don't create several websites for one request. type is linux (PHP, standard Linux hosting) or html (static files only, no PHP or databases). label is an optional name shown in the portal. The website gets a generated domain name, eg. abc123.beeweb.se, with DNS records and an HTTPS certificate set up automatically; custom domains are added in the portal. No database is created. A user can create at most 10 websites per 24 hours with this tool. Creating a website takes a while, usually a minute or two and sometimes longer. The website is returned right away with status provisioning; the file tools can't be used until its status is active. Call wait_for_website to wait until it is ready. If status becomes failed, the website could not be created; tell the user to contact beebyte support instead of creating another one.
create_website
Create a MySQL database for a website with PHP (see list_websites) and write its credentials to the file /database.php in the website's directory, outside the document root /httpdocs. A website can have one database, with one user that has full access to it. If the website already has one, no new database is created: this sets up the existing one and rewrites /database.php, eg. if the file was removed or the password was changed in the portal. Creating a database takes a few seconds, or longer if the website itself is still being created. This waits up to 20 seconds; if status is still provisioning, call it again. config_path is set once the file is written. The password is not returned. The website's PHP code loads the credentials (host, port, database, username, password) from the file, eg. from a file in /httpdocs: $db = require __DIR__ . '/../database.php'; $pdo = new PDO("mysql:host={$db['host']};port={$db['port']};" . "dbname={$db['database']};charset=utf8mb4", $db['username'], $db['password']); The file is replaced each time it is written, so changes to it are lost. Use run_database_query to create tables and look at data. Databases are removed in the portal.
create_website_database
Create a directory on a website, and any missing parent directories. An existing directory is not an error. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
create_website_directory
Create a one-time URL for downloading a directory of a website as a zip archive, eg. /httpdocs for the whole website. Use it to back up a website before replacing or deleting its files, or to work on an existing website's files locally. Needs commands with network access (eg. curl): run the returned command to save the archive. The archive's entries are relative to path, so it can be restored with create_website_upload_url for the same path and overwrite (files added since then are kept, delete them first for an exact restore). Symlinks and files that can't be read are skipped, the X-Archive-Skipped response header says how many. The URL can be used once and is valid for 15 minutes. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
create_website_download_url
Create a one-time URL for uploading a zip archive to a website, which is extracted into the directory path, eg. /httpdocs to publish a website. Use this instead of upload_website_archive when you can run commands with network access (eg. curl), especially for images and other binary files: the archive is uploaded directly, not through tool calls. Zip the files relative to the directory they should end up in (index.html at the top of the archive to get /httpdocs/index.html), then run the returned command with the archive's file name. Archives can be at most 100 MB. The URL can be used once and is valid for 15 minutes. path and overwrite are fixed when it's created. The upload responds with JSON (the number of files and directories extracted and their total size) or an error message. Extraction works like upload_website_archive: the directory and missing parents are created, existing files are only replaced with overwrite, and if the archive is refused nothing is changed. Files that are not in the archive are kept. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
create_website_upload_url
Delete a DNS record of a domain (ids are returned by list_dns_records). Returns the deleted record. Deleting a record can take a website or email offline, confirm with the user before deleting records that are in use.
delete_dns_record
Delete a file, symlink or directory of a website. Non-empty directories are only deleted with recursive, which deletes everything in them. Deleted files can't be restored. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
delete_website_file
Download a file of a website, at most 1 MB. Read larger text files, eg. logs, with read_website_file. Text files (UTF-8) are returned as text with encoding utf-8, other files base64 encoded with encoding base64. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
download_website_file
Get the status of a website's MySQL database, and its name, host, port and username. The password is not returned: create_website_database writes the credentials to /database.php for the website's code (call it again to rewrite the file). If the database is still being created (status provisioning), this waits up to 30 seconds; call it again if it is still provisioning. host and port are set once the database is active. failed means the database could not be created, the user should contact beebyte support.
get_website_database
Get the type, size, permissions and modification time of a file or directory of a website. Fails if it doesn't exist. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
stat_website_file
Get a website's hosting settings: its document root (relative to the website's directory, as used by the file tools), PHP settings and databases. php is null if the website has no PHP; php and document_root are also null for web forwards and websites that are being created or deleted. version is the PHP version, eg. 8.3. mail_function is false if PHP's mail() is disabled, then send email with SMTP instead. databases lists the website's databases, see get_website_database.
get_website_info
List the DNS records of a domain whose DNS beebyte hosts (dns_hosted_by_beebyte in list_domains). Only administrators of the domain's organisation can see and change its records. name is the record's full name (the domain itself for the root, eg. example.com, or eg. www.example.com). data is the record's value: an address for A and AAAA, a hostname for CNAME, ALIAS and NS, "priority mail_server" for MX, "priority weight port target" for SRV, the text for TXT. locked records are managed by beebyte (eg. the SOA and NS records) and can't be changed.
list_dns_records
List the user's domains. Lists the domains in all organisations the user has access to, or only in the given organisation (ids are returned by list_organisations). name is the ASCII form of the domain name, unicode_name its Unicode form for internationalized names. external domains are registered with another registrar and only have their DNS at beebyte; other domains are registered through beebyte and renewed automatically on expiration_date until they are canceled. dns_hosted_by_beebyte is true if the domain uses beebyte's name servers (name_servers), so its DNS records are managed at beebyte. status is active, waiting (see waiting_for: payment of the invoice, or the registrant's BankID validation), registering or transferring (being registered or transferred to beebyte), failed (the registration or transfer failed) or canceled (canceled, the domain expires on expiration_date and is not renewed). Domains are registered, transferred and canceled in the portal (portal_url).
list_domains
List the organisations the user has access to. Includes organisations the user is a member of and organisations the user can access through a reseller. Use the ids to filter other tools by organisation. is_admin is true if the user is an administrator of the organisation. suspended organisations have limited access in the portal, eg. because of unpaid invoices.
list_organisations
List a directory of a website. website_id is the id from list_websites. Paths are relative to the website's directory: the document root (the public web pages) is /httpdocs, subdomains have their own directories, eg. /blog.example.com. Symlinks are listed, not followed.
list_website_files
List the user's websites (webhosting subscriptions). Lists the websites in all organisations the user has access to, or only in the given organisation (ids are returned by list_organisations). type is linux, windows, wordpress, html (static files only, no PHP or databases) or forward (a web forward to forward_url); plus is true for the Plus (high speed) variants. php is true if the website can run PHP, see get_website_info for its version. ipv4_address and ipv6_address are the addresses the website's DNS records should point to. status is provisioning while a new website is being created, see create_website.
list_websites
Move or rename a file or directory within a website. Missing parent directories of the destination are created. The destination is only replaced with overwrite. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
move_website_file
Check the connection to the beebyte portal. Returns the portal user this connection acts as and the organisations the user has access to.
ping
Read lines of a text file of a website, also files too large for download_website_file, eg. logs. Returns lines lines from line start (numbered from 1), or with tail the last lines, eg. the latest entries of a log, at most about 100 000 characters; long lines are cut. start in the result is the number of the first line returned, more says whether there are lines after them. Use it to read around a line from grep_website_files. Files ending in .gz, eg. rotated logs, are decompressed. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
read_website_file
Run SQL in a website's MySQL database (see create_website_database), eg. to create tables, add data or look at what the website stored. sql can be several statements separated by semicolons. They run in order with autocommit, so each change is saved right away, and running stops at the first error: results has one entry per statement that ran, and error has the MySQL error code, message and the number of the statement that failed. Statements that return rows give columns, rows (values as JSON; decimals, dates and times as strings) and row_count; at most max_rows rows are returned in total, truncated is true when some were left out. Other statements give affected_rows and last_insert_id. Each request may run for about 30 seconds. Queries run as the website's database user and can only reach its own database. Changes can't be undone: confirm with the user before dropping tables or deleting or overwriting data they may need, and don't run destructive SQL against a website that is in use without checking first.
run_database_query
Search the contents of a website's text files, like grep: eg. find where a function or text is in the code, or errors in the logs. pattern is a Python regular expression, or plain text with fixed. Searches the file path, or the files below the directory path; glob limits which files by name, eg. *.php. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files). Returns matching lines with their path and line number, and context lines before and after. Returns the first max_matches matches, or with last the last ones, eg. the latest in a log. Files ending in .gz, eg. rotated logs, are decompressed, binary files are skipped. Long lines are cut. truncated means that more lines matched (total) than were returned, and complete false that the search stopped before reading everything: search a smaller path.
grep_website_files
Find files and directories of a website by name. To search their contents use grep_website_files. pattern is a glob matched against names (not paths), eg. *.php or wp-config.php. Searches below path. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
search_website_files
Change a DNS record of a domain (ids are returned by list_dns_records). name, data and ttl work like in add_dns_record; the ones that are left out are kept. The type can't be changed: delete the record and add a new one instead. The old value is replaced and can't be restored by this tool, so note it if the user may want it back. Confirm changes to records that are in use (eg. a website's A records or MX records) with the user.
update_dns_record
Upload a zip archive to a website and extract it into the directory path, eg. /httpdocs to publish a website. content is the zip archive, base64 encoded. Use this rather than upload_website_files for more than a few files. If you can run commands with network access (eg. curl), create_website_upload_url is better, especially for images and other binary files. Zip the files relative to the directory they should end up in (index.html at the top of the archive to get /httpdocs/index.html). The directory and any missing parents are created. Existing files are only replaced with overwrite, other files in the directory are kept. Every entry is checked before anything is written, so if the archive is refused (eg. a file exists, a symlink or a path with ..), nothing is changed. Returns the number of files and directories extracted and their total size. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files).
upload_website_archive
Upload one or more files to a website, at most 100 per call. For more than a few files, upload_website_archive is faster. Each file has a path, content and encoding: utf-8 for text, base64 for binary files. Missing parent directories are created. Existing files are only replaced with overwrite, which keeps their permissions. Each file is written atomically; the files are uploaded one by one and the result says for each whether it succeeded, a failed file doesn't stop the rest. Paths are relative to the website's directory, the document root is /httpdocs (see list_website_files), so to publish a website upload it to /httpdocs.
upload_website_files
Wait for a website that is being created (status provisioning) to become ready, for up to timeout seconds. Returns the website as soon as its status is no longer provisioning, or when the timeout is reached. If the status is still provisioning, call this again; creating a website usually takes a minute or two. active means the website is ready and its files can be managed, failed means it could not be created (the user should contact beebyte support). A new website's domain name can take a few more minutes to resolve everywhere, as DNS caches expire.
wait_for_website
Beebyte ChatGPT Plugin FAQ
How the directory, categories and Discoverability Score work.
Read the methodologyHow do I improve Beebyte's ChatGPT Plugin discoverability?
The levers are the listing surface agents actually read: names, descriptions, keywords, tool metadata, and registry health. Which lever matters depends on where discovery breaks, which is what continuous measurement shows.
What are Beebyte alternatives on ChatGPT?
As of 2026-11-20, Beebyte competes with agent.cloud, AI SPACE, Ample, Apphaven, Aqteron, ASI Cloud, Back4app, BasicDeploy and 27 more in ChatGPT App Hosting & Deployment Platforms, ranked by public Discoverability Score.
Where is this profile measured?
This profile uses the geography attached to the latest public registry snapshot: US. Locale tags are intentionally omitted.