floot
Server Details
Build and publish full-stack web and mobile apps on Floot from any MCP client. Building is free.
TDQS
Scored across 58 tools
Many tools are genuinely distinct, but there are clear clusters of near-overlap: run_code_in_vm vs run_code_in_browser vs query_database vs execute_sql vs call_app_function, and read_file vs read_files vs fetch vs search_code. The long descriptions do a lot of lifting to separate them, but the internal card_* bridges and the upload/create-image/request-upload trio add confusion rather than clarity.
The dominant pattern is consistent snake_case verb_noun (create_project, list_files, delete_branch, execute_sql, publish_app). Deviations are few but noticeable: card_confirm_merge and card_upload_asset put the noun first, and bare verbs like fetch, search and typecheck break the pattern.
58 tools is heavy for any single agent surface and sits in the '25+ feels bloated' band. The domain (full app platform: files, deps, DB, branches, publishing, resources, storage, testing) justifies breadth, but internal UI bridges and low-level escape hatches inflate the count beyond what an agent needs to reason about at once.
Coverage is broad and coherent: file read/write/patch/rename/copy/delete, dependency management, SQL read/write plus schema pulls, branch and merge lifecycle, publish/unpublish/status, resource provisioning, storage, images, tests and logs. Gaps are minor (no project delete/rename, storage listing only via a VM snippet), and agents can work around them.
Available Tools
58 toolsadd_dependencyAdd DependencyAInspect
Add npm packages to the project (validated against Floot's supported set — rejected packages get a supported alternative named; some versions are pinned/substituted). Avoid node-gyp/native packages (exception: sharp is supported, auto-pinned), WASM modules, and packages bundling large binaries (e.g. ffmpeg/ffprobe); pure JS/TS preferred. A bare kysely keeps the version the project already records (a project without one gets 0.28.17), the version its db/schema helpers are written against; pass an explicit kysely@<version> only when upgrading it deliberately. Installs on the project VM and persists resolved versions. After a slow install completes as a job, call add_dependency again with the same packages — the second call is fast and persists.
Inputs: packages (required), projectId (required).
apply_patchApply PatchAInspect
Apply a V4A patch to a Floot project — create and update multiple files in ONE atomic operation. Format: "*** Begin Patch" envelope with "*** Add File: path" (+ prefixed lines) and "*** Update File: path" (hunks: optional "@@ anchor" locator, space-prefixed context, -/+ lines, optional "*** End of File"), then "*** End Patch". Paths follow the Floot item scheme (see read_file). To replace a file wholesale use Add File on its own path; the item's other files (e.g. its .module.css) are kept. This tool does not delete or rename: "*** Delete File:" sections are skipped and listed in the result (delete_file removes files), and "*** Move to:" is rejected (rename_file renames items and rewrites their importers). Every applied patch is recorded in the project history and can be reverted. Keep each patch MODEST (a few files / few hundred lines): chat clients cap per-message output, and a patch cut off mid-way is rejected whole ("missing *** End Patch") — split big changes across several apply_patch calls.
Inputs: patch (required), projectId (required), expected_version.
call_app_functionCall App FunctionADestructiveInspect
Call one function of this project's own backend, the way its own code would: a function the project exports from helpers/flootAppExports.tsx, run in its published app, returning that function's JSON result. list_files shows the project's exported functions with their signatures; the function name and its input come from there. Use it to test an export after publishing it, or to do something for the user through the app's own logic instead of raw SQL. The function runs for real (a function that writes, sends or charges does so), so call one that changes anything only when the user asked for that change. Only functions in the last publish; needs edit access to the project. Up to 25 seconds; no answer in time does not mean it did not run. Details: get_guides("app-calls").
scheduled_job (of static/__dev/scheduled-jobs.json) instead: runs it once now on the preview backend, like the editor's Run now; no publish or branch switch needed.
Inputs: projectId (required), input, function, scheduled_job, +1 more.
cancel_requestCancel RequestADestructiveInspect
Withdraw a pending request you created — a credential request from request_external_resource, a custom-domain setup request from publish_app, or an open screenshot job from screenshot_preview (jobId from that tool). Only pending requests can be cancelled — completed ones are final — and cancelling is final too: a cancelled request can't be resumed (the user would need a new one). Use when the user says to stop or they don't want to proceed.
Inputs: jobId (required), projectId (required).
card_confirm_mergeMerge Card BridgeADestructiveInspect
Internal bridge for the merge card's Merge button (not for agents).
Inputs: jobId (required), secret (required).
card_upload_assetUpload Card BridgeAInspect
Internal bridge for the upload card (not for agents — use request_user_upload / upload_asset). phase 'presign' mints the PUT URL for the picked file; phase 'complete' verifies the object landed and finishes the request_user_upload call with the publicUrl.
Inputs: jobId (required), phase (required), projectId (required), file_name, +2 more.
compare_branchesCompare BranchesARead-onlyInspect
Show what merging one branch into another branch of the same project would change. projectId is always the TARGET (the branch merged into); sourceProjectId is the branch merged in.
Output: changed files, every conflict with its marker text, the migration script, key conflicts, new env vars the target lacks, missing resources, and targetVersion / sourceVersion for start_merge.
With no conflicts it also starts the pre-merge checks in the background. Neither branch is changed.
Conflicts: a merge from x into y edits y; a merge into main edits the source branch (main is never edited by a merge). Fix each conflicted file there with edit_file / apply_patch so it holds both sides' changes.
When to use: the user asks to merge branches, what a merge would change, or to fix a failed merge. Flow: compare_branches → read_merge_conflict and fix each conflict → static/__dev/migration.js if the database or storage must change (get_guides("fork")) → compare_branches → start_merge → get_merge_status.
Inputs: projectId (required), sourceProjectId (required).
compare_published_sourceCompare With Published SourceARead-onlyInspect
Read-only: the source changes a publish now would ship. Current saved code vs the live release (or release): both source digests (equal = identical) and every changed file with both sha256s, paged by offset; diffs only for the paths given. Paid plans.
Inputs: projectId (required), diffs, offset, release.
copy_fileCopy FileAInspect
Copy one or more items to new names (e.g. {from:'components/Card', to:'components/BigCard'}). Item names without extensions; same type only. Importers of the source are left unchanged. Pass several copies to apply them in one call.
Inputs: copies (required), projectId (required), expected_version.
create_checkpointCreate CheckpointAInspect
Create a NAMED checkpoint — a labeled restore point the user sees in the project's Checkpoints panel and can revert to later. All file/dependency changes since the previous checkpoint are grouped under it. Call this AFTER completing a coherent unit of work (a feature, a fix, a requested change set) — not after every file write. Give it a short user-meaningful title describing what was accomplished (e.g. 'Added login page with email auth'), optionally a description with detail. No-op when nothing changed since the last checkpoint. Restoring a checkpoint reverts code and project config only — database rows, uploaded assets and published deployments are not rolled back.
Inputs: title (required), projectId (required), description.
create_projectCreate ProjectAInspect
Create a new Floot project (pre-seeded with the shared component library) and return its id. initial_prompt describes the app to build — what it does and for whom — keeping the user's own wording where they gave one rather than a summary. It becomes the project's system prompt (static/__dev/system-prompt.md, which the user can edit) and grounds the project (served back as in list_files). The result renders a live preview card for the user and includes the first-build playbook: a fresh project is EMPTY until pages are written, so a session normally continues straight into get_guides("design") and the first page rather than ending at the card.
Inputs: name (required), initial_prompt (required).
delete_branchDelete BranchADestructiveIdempotentInspect
Permanently delete a branch: its files, database copy and comments (merged code stays). Only when the user asked. A branch not merged into main needs discardUnmerged: true.
Inputs: projectId (required), discardUnmerged.
delete_fileDelete FileADestructiveInspect
Delete a project file. Deleting an item's main code file (e.g. components/Foo.tsx) removes the whole item including its css/tests; deleting an aux file (e.g. Foo.module.css) only clears that part.
Inputs: path (required), projectId (required), expected_version.
delete_storage_fileDelete Storage FilesADestructiveInspect
Delete files from the project's storage (what the app uploaded or serves from /_cdn/), by their full keys as run_code_in_vm's _floot.storageList returns them ("public/…" or "private/…"), up to 100 per call. The bytes are gone for good. For a public file of a published app, its CDN copy is removed too; the result says what to tell the user about that. Only when the user asked for the files to be deleted: to stop using a file, remove its reference from the code instead. Needs edit access to the project.
Inputs: keys (required), projectId (required).
edit_fileEdit FileAInspect
Replace old_string with new_string in a project file. old_string must match the current content exactly (including whitespace) and be unique unless replace_all is set. Prefer this over write_file for changes to existing files.
Inputs: path (required), projectId (required), new_str, old_str, +4 more.
execute_sqlExecute SQL (Writes)ADestructiveInspect
Run WRITE SQL on the project's Postgres database (Floot-managed or an external one the user connected): CREATE/ALTER, INSERT, UPDATE, DELETE, DROP, migrations. Your MCP client shows the user the SQL to approve. Schema changes re-pull the typed schema helper and return the new schema. Pass database only if the project has more than one. By default the whole query (every statement in it) runs on one connection in one transaction that Floot opens and commits: if any statement fails, or the query times out before its COMMIT, nothing is applied. Leave BEGIN/COMMIT/ROLLBACK out of the query — a COMMIT or ROLLBACK inside it ends that transaction early, so a later failure no longer undoes what came before. Set no_transaction for statements that cannot run inside a transaction block (VACUUM, CREATE INDEX CONCURRENTLY, …); each statement then applies as it finishes. Queries are stopped after 75 seconds either way; split a longer migration into several calls. A disconnect does not stop a call that has started: if you got no result, check the schema before running it again.
Inputs: query (required), projectId (required), database, no_transaction.
fetchFetch Search ResultARead-onlyInspect
Fetch a search result by id: a project overview ('') or a file (':'). For direct access to a known file or project, read_file/list_files give more detail.
Inputs: id (required).
fork_projectFork ProjectAInspect
Fork a project (main or any fork) into a new branch with its own database branch and test keys, when the user asks to fork, branch, or try changes away from the project.
Forks at once; no confirmation.
copyFlootDb (default "full"): "full" the parent's data. "schema": same tables, no rows. "pointInTime" + databaseAt (ISO 8601 with zone): the parent's database as it was then (code, files, keys stay current), to recover older data.
isolated true (only if asked): empty tables; no real data, keys or stored files (the /_cdn/static assets still show).
scratch true: a throwaway branch (QA run, rehearsal, backup copy): cannot be merged, deleted after 7 idle days unless kept.
ref: your own text for the branch (≤200 chars, e.g. a task link); list_branches shows it.
Output: fork project id and preview link. Merge back with compare_branches, then start_merge. Code that breaks on a fork: get_guides("fork").
Inputs: name (required), projectId (required), ref, scratch, +4 more.
generate_imageGenerate ImageAInspect
Generate AI images into the project's asset storage, for PROJECT-SPECIFIC imagery (mascots, illustrations, app icons, a particular style); generate them BEFORE building the components that use them. Returns project-relative paths (/_cdn/static/...) that resolve on the app's own domain: use them in code, or as the app icon via update_project_metadata iconUrl. An image you already hold as a file goes in with upload_asset. To edit or restyle an image, or keep a character, product or logo consistent, pass it in reference_images (this project's stored assets only: upload an outside image or the user's photo first) and say in the prompt what to change.
Inputs: images (required), projectId (required).
get_analyticsGet Visitor AnalyticsARead-onlyInspect
The published app's visitor analytics, as the editor's Analytics tab shows them: visitors, page views and a daily trend, top pages, referrers, countries, devices, browsers and custom events (trackEvent) over a date range. Read-only. Tracking setup and consent: get_guides('analytics').
Inputs: projectId (required), to, from, limit.
get_branch_settingsGet Branch SettingsAInspect
Read a branch's six switches (all off by default): Receive main's email, Send email, Scheduled jobs, Receive main's realtime messages, Send realtime messages to main, Queued jobs. Lists only those the branch's code uses, as the UI does.
When to use: a branch sends or receives no email, its scheduled or queued jobs don't run or run twice, or realtime messages or presence aren't shared with main.
The user changes them in the branch's Project Settings › Advanced (branch creator or project owner). On an isolated branch, scheduled_jobs sets Scheduled jobs; to run one job now, call_app_function's scheduled_job.
Inputs: projectId (required), scheduled_jobs.
get_current_contextGet Current Editor ContextARead-onlyInspect
What the user is looking at RIGHT NOW in their open Floot editor: the active page/component, the preview element they selected (mapped to source file:line), the preview device size, whether they drew a screenshot annotation for you, whether they REVERTED recent changes (undoing edits — re-read before editing if so), and any requests they queued via editor action buttons ("Fix with Floot" etc.). Call this FIRST when the user refers to something without naming it ("this", "here", "the button"), reports a problem without saying where ("it's broken", "looks wrong"), or implies they triggered something in Floot ("go", "I clicked fix", "I undid that"). Cheap and text-only. With several windows open, one answers (the result says which) — but a selection made in ANY window is merged in, so a "no selection" from the answering window plus a deposited selection from another window means the deposited one is what the user means. If it reports a pending annotation, call view_annotation to see the image.
Inputs: projectId (required).
get_guidesFloot GuidesAIdempotentInspect
Floot documentation for agents. Call with no arguments to list available guides. Pass topic for one guide (e.g. topic:'floot-overview') or topics (an array of ids) to fetch several at once. floot-overview explains how Floot projects work — read it before your first code change. Skill guides that ship seed code (marked in the list) AUTO-INJECT it into the project the first time they're loaded with a projectId — pass projectId whenever you're working on a project; idempotent, never overwrites existing files.
Inputs: topic, topics, projectId.
get_job_statusGet Job StatusARead-onlyInspect
Poll a pending tool call by its jobId. Each poll either returns the final result (succeeded/failed/cancelled), or reports the call as still running — call it again until you get the result. Failed calls return their stored error message. A jobId belongs to exactly ONE task: it never blocks other tools or other jobs (run them freely in parallel), and once terminal it is frozen history — a NEW user request means a fresh call on the originating tool, never re-polling an old jobId. Legacy v!/b! job ids are also accepted.
Inputs: jobId (required), wait_seconds.
get_live_paymentsGet Live Payments (Read-Only)ARead-onlyInspect
Look up your Floot Payments records in Stripe across all your projects: one record by id, or a list of one record type. Read-only. Reads LIVE mode unless mode is "test", which reads the test payments made in the preview (those never appear in the user's Stripe dashboard). Refunds, cancels and every other change are not possible here; for those the user logs in at stripe.com. Read get_guides('live-payments') before your first call. Responses follow Stripe API version 2026-08-26.dahlia.
Inputs: object (required), id, mode, page, +9 more.
get_logsGet LogsARead-onlyInspect
Call FIRST when debugging a runtime problem (a 500, a failed request, a blank page, 'it doesn't work'), before theorizing. Sources:
'server' (default): the dev backend's request logs, last hour (method, URL, status, duration, log lines), background jobs included; log_reference_id: one request in full.
'browser': console output and client network requests (
⇄ METHOD url → status, error body of failed ones: CORS, timeouts, third-party calls) from the user's open editor (~1 hour; empty if none); a line'sref=<id>is a 'server' log_reference_id.'background': background work in the preview and the published app: schedules and next run, runs in progress, runs long or cut off at the 15-minute limit in 7 days ('still running?', 'timed out?').
'published': the PUBLISHED app's backend logs (kept 90 days), oldest first, last hour by default; from/to (ISO 8601), filter (literal substring), limit (default 200), cursor (get_guides('prod-backend-logs')).
'push': push sends, newest first, per device, delivered or not; from/to, subscription (one device's endpoint), cursor.
'email': the test inbox: mail kept unsent (branch with Send email off; preview sends to test addresses like qa+1@example.com) and other preview sends, as text with links (sign-in codes, reset links).
'events': project events this app received or sent (onProjectEvent deliveries and their attempts).
Inputs: projectId (required), to, from, limit, +5 more.
get_merge_statusGet Merge StatusARead-onlyInspect
State of a merge by mergeId (merges started in the UI too): running, done or failed. On failed: the failing step or migration script and its exact error. On done: the keys added to the target and the passed pre-merge checks.
Inputs: mergeId (required).
get_preview_urlGet Preview URLARead-onlyInspect
Show the user a live preview card and return the preview link. On an EXISTING project this is typically called EARLY, before the first change, so the user watches edits live from the start; the result is informational and a working session normally continues past it. It is not needed right after create_project, whose result already includes the same card. The preview URL carries an access token in its query string and only works shared EXACTLY as returned (no Floot login, view-only — which is also what lets it open on a phone); it live-updates as you edit, so it also suits an in-app browser tab if your client has one. The result additionally includes the sandbox API base for your own headless /_api/* testing — the frontend does not render there and it is not a user-facing URL; the preview link is the one meant for the user. Floot hosts the app: publish_app publishes it to production. To send the user straight to one view (a page, a component's examples, a page's staged screens, a design canvas or one of its frames, a page's link preview card), pass targetType with path or name, as navigate_preview takes them: the link then opens on that view, with or without a Floot window open.
Inputs: projectId (required), name, path, targetType.
get_publish_statusGet Publish StatusARead-onlyInspect
Read-only publish snapshot for a project: published (true/false, with the live URL when published), customDomains (the user's own domains attached to the project — apex and www are listed separately; empty when none), paid (the workspace owner has a paid plan, which allows removing the Floot badge), plan (free | pro | power — paid cannot tell Pro from Power), nativeBuilds (the owner's remaining monthly iOS/Android app-build allowance; unlimited is true on plans where the ceiling is only a fair-use backstop, and while an Action Boost is live (boostUntil says until when; builds started before then never count toward the allowance) — read this BEFORE passing mobile: true, and on a metered plan build only when the user asked for a native/TestFlight/Play build), displayFlootLogo (whether the live app shows the 'Made with Floot' badge; true until a paid owner turns it off), and mobileBuild (the project's latest native app build: building, succeeded, or failed with what to fix; on a succeeded build, storeBuilds says how each store's build and upload went (building, live, or failed), which can fail after the native build succeeded; isMobileTestMode marks a mobile test build — an app build finishes minutes AFTER the publish job it rode on, so read this to learn how it ended before telling the user it worked). This is the ONLY tool that reports attached custom domains — read it before telling a user whether their domain is connected, and never conclude a domain is unattached from any other output. The publish card calls this on load to render fresh state; also check it before publishing to know whether publish_app will publish fresh or publish the live app again.
Inputs: projectId (required).
list_branchesList BranchesARead-onlyInspect
List a project's main project and every branch you can open, with their ids, names and status. projectId may be the main project or any of its branches.
When to use: the user names a branch of a project, or asks which branches it has.
Inputs: projectId (required).
list_filesList Project FilesARead-onlyInspect
List a Floot project's virtual file tree with sizes, plus its dependencies, current version (pass the version to write tools as expected_version), and current project metadata — title, description, app icon (iconUrl), splash screen, mobile app id, SSR, iOS Info.plist overrides, share target (iOS + Android), native system bars. This is where to look up those settings; update_project_metadata changes them.
Inputs: projectId (required).
list_projectsList ProjectsARead-onlyInspect
List your Floot projects (id, name, last-updated, whether an app icon is set), most recently updated first. name_filter is a case-insensitive substring match on the stored name, which is often not the name the user uses for a project — on a small account a filter that matches nothing returns the whole list instead.
Inputs: limit, name_filter.
list_resourcesList Available ResourcesARead-onlyInspect
List the env vars a project's code can use and the resources behind them: (1) resources CONNECTED to the project — usable as process.env. in endpoint code now; (2) the owner's other account-level credentials — reusable, but not usable in code until connected; (3) everything Floot can add. Call it to learn what env vars exist before writing backend code, and BEFORE provisioning or requesting any credential (the owner may already have the one you need). Pass query (case-insensitive substring over names, descriptions, types, and env var names) to filter when the account has many resources. Read-only. Details: get_guides('resources').
Inputs: projectId (required), query.
provision_resourceProvision ResourceAInspect
Provision a Floot-managed backend resource for the project, fully server-side (Floot mints the secrets), and seed the working code for it. Idempotent: re-running returns the existing resource and skips seed files that exist. Available:
database — A Floot-managed Postgres database (Neon). FLOOT_DATABASE_URL is set for the app.
auth — Email/password + session auth (JWT_SECRET, auto-provisions a database if none). Injects auth pages, endpoints, and helpers.
oauth-login — Sign in with Google via Floot's brokered OAuth (FLOOT_OAUTH). Injects OAuth provider classes, login buttons, helpers.
microsoft-login — Sign in with Microsoft via Floot's brokered login (FLOOT_MICROSOFT_LOGIN). Injects button + auth endpoints.
google-integration — Built-in Google integration (FLOOT_GOOGLE_INTEGRATIONS): end users connect their Google account so the app reads/writes their Gmail, Calendar or Drive, unless the user wants their own Google client. See get_guides("floot-provided-google-integrations").
microsoft-integration — Microsoft Graph access (Outlook/Teams/etc.) via Floot's brokered Microsoft OAuth (FLOOT_MICROSOFT_INTEGRATIONS). Injects Connect button + endpoints.
push-notifications — Web + native push (FLOOT_PUSH). Mints VAPID keys, injects helpers/pushClient (subscribe/unsubscribe) + a service worker.
self-edit — Lets the app edit ITS OWN source through the Floot MCP server (FLOOT_SELF_EDIT): mints a project-scoped bearer token, exposed to the backend as FLOOT_PROJECT_TOKEN + FLOOT_PROJECT_ID. Duplicates get their own token. No seed code — see get_guides("floot-self-edit").
payments — Built-in payments (FLOOT_PAYMENTS), unless the user wants their own Stripe account. See get_guides("floot-payments").
shipping — Shipping physical goods through the owner's own Shippo account (FLOOT_SHIPPING: rates, labels, tracking); the preview uses built-in test shipping. See get_guides("floot-shipping"). Other enum values are beta-gated. Sending email needs no resource (@floot/email, get_guides("email")). The user's OWN key (their OpenAI key, an external database) goes through request_external_resource. Turning Floot Payments OFF (publishing is refused while it is attached with Stripe setup unfinished, also when the app no longer uses it): action "turn_off" with resource "payments". Owner only; ask the user first. It keeps the app's data and Stripe setup; the owner turns it back on in Floot (Services → Payments). Turning Floot Shipping OFF (publishing is refused while it is attached and Shippo is not connected, also when the app ships nothing): action "turn_off" with resource "shipping". Edit access; ask the user first. It keeps the app's data and Shippo connection; provisioning "shipping" again turns it back on.
app-connection — let this project's backend call functions ANOTHER Floot app exports (no API keys): pass app_target and app_functions. Only apps the user can edit, only functions in the other app's helpers/flootAppExports.tsx (saved is enough: write both apps first, connect, then publish both); the user confirms in a dialog. Read get_guides("app-calls") first.
Inputs: resource (required), projectId (required), action, app_note, +2 more.
publish_appPublish AppADestructiveInspect
Publish the app to production: the first publish, a new publish of changes the user wants live, or a custom domain. Read get_guides('publishing') before your first call. A live app publishes at once, nothing to pass or confirm. A first publish (no domain) shows the user the publish form, and nothing publishes until they press Publish. Always publishes live (web, the mobile app when built, payments in Stripe live mode): never ask the user test or live; mobile test builds are started by the user on floot.com. If the owner has not finished Stripe setup (payments) or connected Shippo (shipping), the publish is refused with a card to do it (nothing published): tell the owner to click it, then call publish_app again.
Inputs: projectId (required), domain, mobile, add_another, +2 more.
pull_database_schemaPull Database SchemaAInspect
Introspect the database and write a typed schema helper the app uses for queries (kysely on current projects; some legacy projects use drizzle or snake_case kysely — the pull matches whatever the project already uses). Usually NOT needed after execute_sql — schema-changing statements re-pull automatically. Use it to refresh manually, or with helper_name to generate the helper for an additional/external database. The helper is GENERATED — never hand-edit it or cast around its types: if a column's type is too loose (e.g. role as string when code expects "user" | "admin"), fix the DATABASE (CREATE TYPE … AS ENUM + ALTER COLUMN … TYPE) and re-pull, and the union type falls out.
Inputs: projectId (required), database, helper_name.
query_databaseQuery Database (Read-Only)ARead-onlyInspect
Run a READ-ONLY SQL query against the project's Postgres database (SELECT, EXPLAIN, etc.) — the Floot-managed database, or an external Postgres the user connected to the project. Writes are rejected — use execute_sql for those. Returns JSON: {rows, rowCount, command, truncated?} (or {results: [...]} for multi-statement queries). Pass database only if the project has more than one.
Inputs: query (required), projectId (required), database.
read_fileRead FileARead-onlyInspect
Read a file from a Floot project (cat -n style); offset/limit for large files, column to start mid-line (a capped read names the offset and column to continue from). Paths: components/Name.tsx, components/Name.module.css, helpers/Name.tsx, pages/name.tsx, pages/name.pageLayout.tsx, endpoints/route_POST.ts, endpoints/route_POST.schema.ts, static/file.txt, base.css. include_references:true also lists the files that reference this one (imports, queueTask/runCode names, endpoint URL strings): check it before renaming or deleting, or use rename_file, which rewrites importers. Hosted assets read too (/_cdn/ as upload_asset / generate_image return it, private/): png/jpeg/gif/webp come back as an image you can see (≤3.75 MB), text as text, other binaries as a summary. No typecheck by default: diagnostics:"auto" appends a .ts/.tsx file's current type errors when the VM is warm, "wait" boots it and forces the check. older_than walks the file's history: "current" shows it just before its latest change, the returned cursor steps further back; restore by writing the old content with write_file.
Inputs: path (required), projectId (required), limit, column, +4 more.
read_filesRead FilesARead-onlyInspect
Read up to 20 files in ONE call, much cheaper than repeated read_file: prefer it whenever you need several files together (an endpoint, its .schema.ts and the hook that calls it). Same paths and options as read_file (diagnostics, include_references; /_cdn/ images come back as images), each file cat -n style under a header. Output is capped: whole files at the end are left out and listed by name. With older_than (up to 5 paths) the files are walked back through history TOGETHER on one timeline — the way to keep a component and its stylesheet consistent.
Inputs: paths (required), projectId (required), limit, older_than, +2 more.
read_merge_conflictRead Merge ConflictARead-onlyInspect
Read one conflicted file of a merge in pages (cat -n style, 400 lines by default), with its conflict block line ranges. projectId is always the TARGET (the branch merged into); sourceProjectId is the branch merged in.
view: "markers" (the file with conflict markers; default), "source" or "target" (that branch's current file).
Inputs: path (required), projectId (required), sourceProjectId (required), view, +2 more.
remove_dependencyRemove DependencyAInspect
Remove npm packages from a Floot project's dependency record (record-only; nothing runs).
Inputs: packages (required), projectId (required).
rename_fileRename FileAInspect
Rename one or more items and automatically rewrite every file that imports them. Use item names WITHOUT extensions (e.g. {from:'components/OldName', to:'components/NewName'}). Preferred over delete+create — preserves content and fixes importers. Same type only. Pass several renames to apply them atomically in ONE pass; importer rewrites are resolved across the whole batch (including chains where one rename's target is another's source).
Inputs: renames (required), projectId (required), expected_version.
request_external_resourceRequest External ResourceAInspect
Request the user's own credential for an outside service on this project: their OpenAI or Anthropic key, an external Postgres database (type POSTGRES), or any other service's key (type GENERIC, with secret_env_vars). Floot-managed resources (database, auth, push, …) need no user input: use provision_resource. Request it early and keep building; code can reference process.env. before the user connects. Never ask the user to paste a secret into the chat.
Inputs: type (required), projectId (required), name, instructions, +3 more.
request_user_uploadRequest a File From the UserAInspect
Show the user an inline upload card so they can hand you a file from their device (image/font/audio/…) — it lands in the project's hosted assets and the card gives you the hosted publicUrl. This is the path for any file the user has: an image they attached in this chat (attachments never reach MCP servers — you see them through vision only, so the user re-picks the same file here), a file on their machine, or a user-provided file you hold but can't upload yourself (over the 3 MB inline cap with no S3 egress — the card uploads from their browser, which is never egress-blocked). Returns a jobId — poll get_job_status; it stays running until they upload, then returns the publicUrl to reference in code. When you show the card, tell the user in a sentence why you're asking — e.g. that you can see their image but the file itself doesn't reach Floot, so re-adding it here is a one-click step — and ask them to say "uploaded" when done in case your polling ends before they finish. For files you hold yourself, use upload_asset; for AI-generated imagery, use generate_image. Some clients can't show the card (Microsoft Copilot Studio, for one): if the user says they don't see it, give them the project's preview link (get_preview_url) and ask them to sign in there and drag the file into Storage › Static (or use its Upload button), then call list_files — the file is listed under "Files uploaded in the editor" with the /_cdn/static/ path to use.
Inputs: projectId (required), description.
run_code_in_browserRun Code in BrowserADestructiveInspect
Run a TypeScript snippet inside the running app's preview document in the user's open Floot editor and return the string it returns. document and window are the app's own DOM: measure elements, read computed styles and rendered output, or click a button and read the result. Runs at the project root, so project files import by relative path, and browser APIs (canvas, createImageBitmap, WebGL) work. A _floot global has storageUpload({ filename, contentType, sizeBytes }) to save a file the page made (e.g. a canvas as a PNG: it returns a presignedUrl to PUT the bytes to and the file's url), storageGetUrl(key) and getFileById(id) to load a stored or uploaded file into the page, and navigate(path) to move the preview to another route first. Needs the project open in the editor (fails fast otherwise); prefer run_code_in_vm when the DOM is not needed. A snippet gets 20 seconds, fixed (there is no timeout argument); past that the call fails while anything the snippet started keeps running. So never wait on a slow /_api/* endpoint here: call it from run_code_in_vm (timeout_seconds up to 120), and for longer work have the endpoint start a background task and poll (get_guides("queueTask")). Typed text input and long multi-step flows are unreliable here: leave those to the user.
Inputs: code (required), projectId (required).
run_code_in_vmRun Code in VMADestructiveInspect
Run a Node.js snippet on the project's compute VM and return its stdout and stderr. ESM by default (a bare require() snippet runs as CJS). The project's npm dependencies are importable and the network works, so a snippet can call the project's /_api/* endpoints (get_preview_url → apiBaseUrl). The project's env vars, secrets and source files are not available.
For data too large to pass through the conversation (an uploaded CSV, images, usage and log history), a _floot global works on this project only:
runSQLQuery({ query, reasonAndExplanationForNotReadOnly? }): runs SQL on the project's database. Read-only unless a reason is given; then the whole query is one transaction, and DELETE/UPDATE/TRUNCATE/DROP are refused. A call already sent is not cancelled when the snippet times out (guide vm-sql).getFileById(id): the download URL of a file the user uploaded.storageList(prefix),storageGetFileSizes(visibility, continuationToken?),storageGetUrl(key): the project's stored files, their sizes, and one file's download URL.storageUpload({ filename, contentType, sizeBytes }): an upload URL for a new public file, and the URL it will have.getHostingUsage({ days? }),getLambdaUsage({ days? }): the project's daily hosting and compute usage.getProdBackendLogs({ filter?, from?, to?, nextToken?, limit? }),getPushHistory({ subscription?, from?, to?, offset? }): the published app's logs and push sends, to count or search in code.paymentsTestRequest({ method, path, params? }): a Stripe request as the preview's stripeRequest makes it (test mode, project owner only; guide live-payments). For a one-off read use query_database or get_logs; for other SQL, execute_sql; to delete files, delete_storage_file; to run one of the app's exported functions, call_app_function. Guides: vm-sql, static-storage, hosting-cost-analysis, prod-backend-logs.
Inputs: code (required), projectId (required), timeout_seconds.
run_testsRun TestsARead-onlyInspect
Run the project's Jasmine specs (helpers/.spec.tsx) on the project VM (jsdom): specs may render components (@testing-library/react) or call endpoint handlers and backend helpers directly. Limits: fetch throws inside tests (mock with spyOn(globalThis, "fetch")), process.env secrets are absent (give a module that reads one at load a dummy value from a helper the spec imports BEFORE it, e.g. helpers/testEnv.tsx setting process.env.JWT_SECRET = "test-secret"; env is reset per spec file), spyOn cannot replace an exported function of a project module (pass dependencies in as arguments instead), no top-level await in specs, and @floot/ services (email, storage, ai, queueTask, …) never run: each function is a Jasmine spy that throws until the spec gives it a result on the imported function itself, e.g. (sendEmail as jasmine.Spy).and.resolveTo({ ok: true, messageId: "m1" }); spies reset before every spec. Returns per-file PASS/FAIL with failing expectations. Defaults to all spec files except hook specs (file name contains "use" — those need real React scheduling and are excluded, matching the in-editor checker); pass paths to run specific spec files, including hook specs.
Inputs: projectId (required), paths.
screenshot_previewScreenshot the User's PreviewARead-onlyInspect
Capture a screenshot of the user app. Call it whenever you want to SEE what the app currently looks like (layout, styling, rendered state) or want to debug the app. By default it shows whatever device the user has the preview on. Pass viewport to see it on another device — this is how to check a mobile or tablet layout: the preview switches to that device for the capture only, then back. Never build a separate "mobile" page or example just to look at a narrow layout.
Inputs: projectId (required), viewport.
searchSearch ProjectsARead-onlyInspect
Search your Floot projects and their code. Returns result ids usable with fetch. For richer options, list_projects enumerates projects and search_code does code-level search.
Inputs: query (required).
search_codeSearch CodeARead-onlyInspect
Search a Floot project's files (string or regex) with optional glob filters (e.g. ['components/*', 'endpoints/**']). Returns file:line excerpts plus filename matches; capped at 40 results.
Inputs: query (required), projectId (required), glob, regex.
start_mergeStart MergeADestructiveInspect
Merge sourceProjectId into projectId (the target). This changes the target: its code becomes the merged code, and the migration script runs on its database and storage. Only when the user asked to merge.
targetVersion / sourceVersion: from the latest compare_branches. conflictSides: conflict path → "source" | "target" (that branch's current file wins). addKeys: new env vars from compare to add to the target. keyChoices: env var → "target" | "source" for every key conflict ("source" overwrites the target's value). appIcon: take the source's app icon (default: as compare says).
Waits up to ~40 s for the pre-merge checks ("Checks still running" → call again), then starts the merge and returns its mergeId for get_merge_status.
A started merge ends done (code + migration applied) or failed (nothing changed on the target).
Into a protected main: not started; the user presses Merge in the card it shows, then get_job_status gives the mergeId.
Refused, with the reason, on failed checks, unresolved conflicts, changed versions, an already merged branch or a running merge.
Migration script: static/__dev/migration.js on the branch the merge edits (never main); dry-run, then run in one transaction on the target on every merge, so idempotent; ctx.sourceQuery / targetQuery / storageCopy / storageDelete only, no network, ~7 min per run. get_guides("fork").
Inputs: projectId (required), sourceVersion (required), targetVersion (required), sourceProjectId (required), +5 more.
typecheckTypecheck ProjectARead-onlyInspect
Typecheck the project (incremental tsc on the project VM). Type errors don't block the app from running. Without paths it checks only code that ships: spec (.spec.tsx), example (.example.tsx) and design (designs/) files are skipped, and the result says how many. Pass paths to check named files, those included, plus their direct importers. run_tests runs specs.
Inputs: projectId (required), paths.
unpublish_appUnpublish AppADestructiveInspect
Take the published app offline and release its subdomain — destructive, confirm with the user first. Details: get_guides('publishing').
Inputs: projectId (required).
update_project_metadataUpdate Project MetadataAInspect
Update project settings; current values appear at the top of list_files. Changes reach the published app at the next publish. Settings marked "on request" change only when the user asks about that subject.
updates (flat keys): title (2-100 chars), description, iconUrl (app icon: favicon, home-screen and native icon; a /_cdn/static/… path from upload_asset, generate_image or request_user_upload, or an https URL; null resets), splashUrl (native splash, same forms), mobileAppId, enableSSR, flootAiDisallowed (true opts out of @floot/ai), iOS Info.plist NS…UsageDescription strings (null removes) and UIViewControllerBasedStatusBarAppearance (boolean, null restores). Invalid keys are reported and skipped. After the first iOS / Android publish, edit static/__dev/native/ios-info.plist or android-manifest.xml instead (get_guides("ios-info-plist"), get_guides("android-manifest")).
shareTarget: native share-sheet target; ship it with the handler in app code (get_guides("share-target")).
nativeSystemBars: 'inset' (default) or 'edge-to-edge', which needs safe-area padding in the app (get_guides("native-system-bars")). ios.scrollEdgeEffect 'hidden' removes the iOS 26+ blur under the status bar.
serverMemoryMb: expert setting, only when the user explicitly asks to change the server memory: too low stops the backend and can cost more (get_guides("server-memory")).
analyticsMode: on request (analytics, cookies, consent; get_guides("analytics")).
iosDeviceFamily: before 'iphone' or 'ipad', ask whether the app is live on the App Store supporting the device it would drop; if it is, it cannot drop it. The only way to make it iPhone-only or iPad-only.
homeScreenName: the name under the installed app's icon (iOS + Android), also after the first mobile publish; keeps the bundle id and store listing names. null = default (store listing title, else project name).
routePrefetch: on request (page prefetching, load performance; get_guides("route-prefetch")).
dependencyOverrides: on request, to fix a vulnerable or conflicting nested package (get_guides("dependency-overrides")).
securityHeaders: on request (iframes, CSP, a security scan, referrer; get_guides("security-headers")). Never in the preview.
Inputs: projectId (required), updates, shareTarget, analyticsMode, +7 more.
upload_assetUpload AssetAInspect
Upload a binary asset (image, font, audio, …) you actually hold (generated, downloaded or read yourself) to the project's storage. The user's chat attachments never reach MCP servers (you see attached images by vision only): use request_user_upload, where the user picks the file again. For fresh AI images use generate_image.
ChatGPT conversation files (a generated image, a file ChatGPT holds): pass it as
file; the host attaches a download link and this server fetches the full-quality bytes (content_type and size_bytes optional). Never downscale or re-encode to fit the inline cap.Up to 3 MB you hold: content_base64 + size_bytes (decoded count); completes in this call and returns publicUrl.
Larger (max 100 MB): size_bytes alone returns an uploadUrl; PUT the raw bytes with the same content_type and exact byte count (curl -X PUT -H 'Content-Type: image/png' --data-binary @file.png ''), then use publicUrl. Some sandboxes (claude.ai Cowork, ChatGPT containers) block S3: if the PUT fails in any way (connection failure, proxy error, or no x-amz-request-id header in the response) the block is permanent for the session, so switch instead of retrying or shrinking:
filein ChatGPT, content_base64 under 3 MB, request_user_upload for the user's files, or a PUT from run_code_in_vm (mint a new URL first; it is short-lived).
Inputs: file_name (required), projectId (required), file, size_bytes, +2 more.
view_annotationView AnnotationARead-onlyInspect
View a screenshot annotation the user drew on the app preview (annotationId comes from get_current_context). Returns the annotated image — the user's drawings/text point at what they mean. Annotations expire after ~1 day.
Inputs: projectId (required), annotationId (required).
write_fileWrite FileAInspect
Create or fully overwrite a file in a Floot project. Content is written literally. Paths must follow the item scheme (see read_file); invalid paths are rejected with the rule they broke. Pass expected_version (from list_files/read_file) to detect concurrent edits. Writing components/Name.module.css sets the css of components/Name — other properties of the item are preserved.
Inputs: path (required), content (required), projectId (required), expected_version.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Build, deploy, and host full-stack web apps from any MCP client. DB, auth, storage, cron included.
Create, edit, preview, and deploy full-stack web apps to a live URL from any MCP client.
Build, deploy, and operate hosted web apps on VibeKit (vibekit.bot) from any MCP client.
Build and publish full-stack apps from your coding agent: models, rules, pages, auth, per-app MCP.
Related MCP Servers
AlicenseNot gradedqualityBmaintenanceFloom MCP server — deploy Python apps instantly with auto-generated UI, REST API, and shareable URLs. 32 tools for AI agents.47MIT- AlicenseAqualityDmaintenanceMCP server for the FloopFloop API that lets LLM hosts build, poll, and refine projects from natural language prompts.1211 npmMIT
- AlicenseAqualityCmaintenanceBuild and host full-stack web apps and sites on Hatchable from any MCP client. Each project gets a dedicated Postgres database, auth, storage, custom domains, and cron.332MIT
- FlicenseNot gradedqualityDmaintenanceAn MCP server that enables Claude to deploy full-stack web apps to Cloudflare, including databases, authentication, and file storage, directly through natural language.-
Glama MCP Gateway
Add one secure layer between your agents and this server.