VibeDeploy

Discard a begin_deploy session and its staged files; the live site is untouched. Cleanup is best effort and retried later if it fails.

Start attaching a user-owned domain to an existing site: returns the DNS records the user adds at their DNS provider (a routing record that sends visitors here, plus a TXT that proves ownership). Idempotent: the same siteName and domain return the existing record. Once the records are published (usually minutes, up to 24h), call verify_custom_domain with the returned recordId; VibeDeploy also re-checks automatically for a few days and emails the user when the domain is live. The site must already exist; call deploy_site first if it does not.

Send one file in several calls when it is too big for a single tool call. Last resort: first try add_files with encoding:'gzip+base64' (about 250 KB of text in one call), begin_deploy's uploadUrl (one 100 MB tar POST) or deploy_from_url (public archive URL). When chunking, gzip+base64 each chunk too, mark the first with isFirst and the last with isLast, and send chunks of one file one after another: parallel chunks corrupt it. After a transport error, do not resend the chunk blindly: restart the file from its first chunk (isFirst truncates it).

Add files to an open begin_deploy session; call it as often as needed, then commit_deploy publishes them together. Paths, extensions and encodings are checked on every call. A 100-250 KB text file usually fits in one call with encoding:'gzip+base64' (gzip locally, then base64); prefer that over add_file_chunk. If your shell can reach mcp.vibedeploy.be, one tar(.gz) POST to begin_deploy's uploadUrl is faster still.

Apply literal find/replace edits to several files in one call, in the served files (target 'dist', live at once) or the editable source (target 'source', shipped by build_and_deploy). Same edit rules as update_file_content. All or nothing: if any edit's match count differs from its expected count, no file is written.

Open a staging session for a deploy that is too large for one call: stage files with add_files (repeatable), add_file_chunk, or one tar(.gz) POST to the returned uploadUrl, then publish with commit_deploy. Use begin_deploy only when the files do not fit in one deploy_site or update_site call. Up to 5 open sessions per token; a session expires after 1 hour.

Build the site's stored editable source in an isolated one-shot build pod and atomically publish the output as the live site, replacing the served files (a restore point is normally taken first). Use build_and_deploy after changing stored source (write_source_files, apply_edits with target 'source'); use update_site instead to upload files that are already built. The first build saves buildCommand, outputDir and rootPath on the site, so later calls can omit them. A failed build returns its log in the error details.

Check whether a domain can be registered and get an indicative price (VAT included where it applies). Read-only: it never buys, registers or reserves a domain. alternatives:true also checks the same name on be, com, net, eu, nl, io, dev and app. Not tied to a site. The result states whether the user's plan includes buying domains in the dashboard (purchase.dashboardUrl); a domain the user already owns is connected with add_custom_domain.

Publish a begin_deploy session's staged files to the live site in one atomic step, after a secret and malware scan of the complete set (on failure the session stays open to retry or abort). Replace mode on a site that has files needs confirm:"I-want-to-replace-all-files". After a timeout or dropped connection, call it again with the same deployId: it returns the running or finished commit instead of starting a second one. Use commit_deploy only to finish a begin_deploy session; deploy_site and update_site publish in one call.

Take a point-in-time snapshot of a site's served files, a restore point the user can restore in the VibeDeploy dashboard within the plan's restore window. Served content does not change, but retention cleanup may remove older snapshots. Container data is not included. Requires an owner or admin team role. NO_VOLUME means the site was never deployed.

Soft-delete a site (confirm:true required). It stops serving at once and is purged after the plan's restore window (purgesAt in the result); until then the user can restore it in the VibeDeploy dashboard. Use delete_site to remove the whole site; use remove_app instead to stop only its container services.

Remove one file from the site's editable source tree. The served files do not change until build_and_deploy runs.

Run container services next to a site, from any public image: the user's own server (running code from the site's source tree), a database such as Postgres or MySQL, Redis, or workers, each in a gVisor sandbox. The call applies the whole config: services missing from it are stopped (their volumes are kept) and the others restart, picking up new source files. It waits until every service is ready and returns per-service status; if one is not ready, call get_app_status with logs. One service may be public: it serves the site's HTTP traffic on its port. Use "${generated:NAME}" for database passwords. API keys and other outside secrets never go in the config, the source or the chat: the user adds them in the VibeDeploy dashboard (site Settings, Secrets) and every service receives them as environment variables. Per site: up to 6 services, 2048Mi memory and 2000m CPU in total, 4 volumes, 3Gi storage (VibeDeploy can raise the CPU / memory ceiling of a site; get_app_status shows it as resourceCeilings). Container volumes are not included in snapshots yet.

Fetch a public tar(.gz) archive over HTTPS and deploy its files to an existing site in one call (begin_deploy, upload and commit_deploy combined). Use deploy_from_url when the files are already published at a public URL, such as a repository archive that holds only the built site (no dotfiles), or when your sandbox reaches the allowed hosts (listed under archiveUrl) but not mcp.vibedeploy.be; otherwise use deploy_site or update_site. In replace mode one top-level folder that wraps every file (as in repository archives) is removed, and a site that already has files needs confirm:"I-want-to-replace-all-files". A failed call never leaves a staging session open.

Publish a static site or single-page app to a live HTTPS URL on the account's platform subdomain (e.g. {name}.subsite.site). A new name creates the site. An existing name REPLACES EVERY FILE (files not in this call are deleted) and needs confirm:"I-want-to-replace-all-files"; preview with dryRun:true. Use update_site instead to change some files of an existing site. A 100-250 KB text file fits in this call with encoding:'gzip+base64' (gzip locally, then base64); use begin_deploy only when the files do not fit in one call. Ship a top-level 404.html for multi-page sites (otherwise unknown paths serve index.html) and use plain ASCII file names. Afterwards add_custom_domain can serve the site on the user's own domain.

Return the team's plan, its limits, current usage, the trial end date and the per-token MCP rate limit. Call it before deploy_site or add_custom_domain to check a plan limit instead of running into PLAN_LIMIT_EXCEEDED.

Show the site's container services: per-service status (ready, restarts, waiting or crash reasons, volumes), the stored config, the names of the site secrets that are set (never values) with the dashboard link where the user adds them, the CPU / memory ceilings of this site (resourceCeilings), and optionally the recent logs of one service. Use get_app_status instead of get_site for anything about containers.

Read the form-to-email relay config of a site, plus the resolved delivery mode, the active From address, and (for a custom sender domain) the DNS records to publish and their verification status. The submission format for the site's form is described under set_forms_config.

Return one site's details: live URL, status, plan, served file paths, whether editable source is stored (and its build settings), custom domains, bandwidth and the last 10 deploys. Use get_site to inspect a site before editing it; use list_sites to find site names, and get_app_status for container services.

Return a privacy-safe traffic summary for a site over the last `period` days (default 7): total page views, distinct-visitor count, top pages, daily counts, device/browser breakdowns, and Web Vitals averages. Never exposes raw visitor IPs or user-agents.

Return all custom domains attached to a site, both those added with add_custom_domain (source 'mcp', with a recordId for verify_custom_domain or remove_custom_domain) and those linked in the VibeDeploy dashboard (source 'dashboard', verified or removed on the dashboard Web Addresses page at manageUrl). Call it before add_custom_domain to avoid adding a domain twice.

List the team's begin_deploy staging sessions, by default the active ones (open and committing), up to 50. Useful after a crash or restart to commit or abort leftovers.

Read the DNS records VibeDeploy tracks for a site (host, type, value), oldest first, including custom-domain verification records. Read-only: it never creates, changes or deletes a DNS record.

Return SHA-256 and size of every served file. Before a redeploy, hash your local files, compare, and upload only the changed ones with update_site mode:'patch' or begin_deploy plus add_files.

Return the site's last 50 deploy and snapshot entries, newest first: how each was triggered, label, snapshot name, file count and secrets detected. Each entry's snapshot holds the content from BEFORE that entry. Restoring is done by the user in the VibeDeploy dashboard within the plan's restore window; there is no restore tool.

List the team's sites (name, URL, status, last deploy), most recently updated first, up to 100. Use list_sites to find a site name; use get_site for one site's files, source status and history.

List the snapshots of a site's served files (point-in-time restore points). NO_VOLUME means the site was never deployed. Restoring a snapshot is done by the user in the VibeDeploy dashboard within the plan's restore window.

List path, SHA-256 and size of every file in the site's editable source tree (the stored pre-build code, not the served files). Read-only. A site without a source tree returns SOURCE_NOT_STORED: write the source with write_source_files, or copy plain static served files with promote_dist_to_source.

Create the editable source tree of a site that has none by copying the files it serves, so later changes can use write_source_files, apply_edits(target:'source') and build_and_deploy. The served files do not change, and a site that already has source is left as is. Use mode 'force' only after telling the user that build output (minified bundles) is not original source.

Return one file the site serves now, as text (utf8) or base64 for binary files, e.g. to edit it and send it back with update_site mode:'patch'. Use read_file for the served files; use read_source_file instead for the editable source tree (get_site.sourceStored). Max 5242880 bytes; get_site.filePaths lists the paths.

Read up to 50 served files in one call, with the same per-file rules as read_file and 8388608 bytes in total; paths past the budget fail with BATCH_BUDGET_EXCEEDED, so request them again in another call. Use read_source_files instead for the editable source tree.

Return one file from the site's editable source tree (the stored pre-build code), so any chat can edit the site without the original local files. list_source_files shows the paths. Use read_source_file for source; use read_file instead for the files the site serves.

Read up to 50 files from the editable source tree in one call (8388608 bytes in total). Each path succeeds or fails on its own; paths past the budget fail with BATCH_BUDGET_EXCEEDED. Use read_source_files for source; use read_files instead for the files the site serves.

Download the external images a site references (e.g. from the old site it was rebuilt from), store them on this VibeDeploy site under assets/img/, and rewrite the HTML <img> references to local paths so the site no longer depends on the original. Call this once AFTER deploying a site rebuilt with the Website Converter. Auto-detects the external image URLs from the site's own HTML; downloads are SSRF-guarded, size/count/time capped, and applied atomically (patch mode). Images already on a VibeDeploy platform domain are skipped.

Stop every container service of the site; the site serves its static files again. Volumes (database data) are kept unless deleteData is true, which deletes them for good. Use remove_app to stop only the containers; use delete_site to remove the whole site.

Detach a custom domain from a site. The site itself is unaffected and keeps serving on its platform subdomain (e.g. {name}.subsite.site).

Search every file of a site for a literal string or a regular expression: the served files by default, or the editable source tree with target 'source'. Use it before reading files in bulk, to find the ones that matter.

Configure the built-in form-to-email relay, fully self-service. Supports a custom From (via a verified sender domain or your own SMTP relay), an explicit Reply-To, and full email branding (subject template, field labels/order, logo, accent color, or a custom HTML body). Requires team role owner or admin. Pass config:null to switch the relay off. If you set a custom `sender` without an `smtpRelay`, the response returns the DNS records to publish; then call verify_forms_sender_domain. Submissions: POST JSON to the returned `endpoint` with Content-Type: application/json (UTF-8). Flat object of form fields (strings/numbers/booleans; checkbox groups may be arrays of strings, joined with ', '). Max 30 fields, 5000 chars/field, 20000 total. Response: {success:true,data:{ok:true}} or {success:false,error:{code,message}}. Rate limit: 10 submits per IP per 10 minutes. Include a hidden honeypot input (default "_gotcha") and leave it empty.

Apply literal find/replace edits to one file in one call, for small changes where uploading the whole file is wasteful. Each edit states how many matches it expects; any mismatch aborts the call with no writes. Served-file (dist) edits go live at once; source edits ship with build_and_deploy. Use apply_edits for several files.

Change files on an existing site. Default mode 'patch' writes only the listed files and keeps the rest; `delete: [paths]` removes files. mode:'replace' deletes every file not listed. Preview with dryRun:true. Use deploy_site instead to create a new site. A 100-250 KB text file fits in one call with encoding:'gzip+base64' (gzip locally, then base64); prefer that over begin_deploy plus add_file_chunk. Fails with SITE_NOT_FOUND when the site does not exist.

Check the records from add_custom_domain (the ownership TXT and the routing record) and, when both are right, start serving the site on that domain; the HTTPS certificate follows automatically, usually within a minute or two. If it fails, the result names each record to add, change or remove; DNS changes can take a few minutes to propagate. Calling it again on a verified domain is a harmless re-check.

Check the DNS records for a site's custom sender domain (DKIM TXT + SPF include). Once the DKIM record is observed, the sender domain is marked verified and the relay sends from the custom From (DKIM-signed). Until then it falls back to the platform address. DNS can take a few minutes to propagate; re-run if it fails the first time. Not needed when the site uses a custom smtpRelay.

Write or overwrite files in the site's editable source tree (the first write creates it). The served files do not change until build_and_deploy runs. Use list_source_files first to see what exists. A legacy site without source storage returns SOURCE_STORAGE_NOT_ENABLED; the user can ask VibeDeploy support to enable it.