MCP HeartbeatTHE MCP OBSERVATORY
Public workspace

Explore

Evidence for your next connection.

Independent measurements of remote MCP servers. Every observation has a timestamp.

Read-only probes

MCP server · Communication & messaging

Dreamlit

Create, test, publish, and manage Dreamlit notification workflows from AI clients.

Checks passed
12 / 22

3 failed · 7 not measured

Median response
796 ms

Median of successful probes over 7 days

30-day uptime
Not measured

Share of countable probes over 30 days, including daily rollups. Auth challenges are excluded.

Discovered tools
16

Tool definitions returned by this server

Independent monitoring

Health & checks

Status history

Responded

The server completed its last measured handshake. This is an observation, not a guarantee.

Last checked
Check summary by test group
Test areaPassedFailedUnmeasured
ConnectionCan a client connect and discover tools?
500
Reliability & speedDoes it respond consistently and quickly?
004
Access & authenticationWhat is required to use this server?
102
Tool qualityAre the tool definitions usable by an agent?
400
Publisher identityCan the publisher be traced to its source?
231

Unmeasured means there is not enough evidence to decide. How these checks work

Inspect individual checks 22 recorded

Core

  • Pass
    Reachable

    Latest probe returned HTTP 200 in 796 ms.

  • Pass
    Valid handshake

    protocolVersion 2025-11-25, serverInfo.name Dreamlit MCP.

  • Pass
    Lists tools

    tools/list returned 16 tools.

  • Pass
    Current protocol

    Negotiated 2025-11-25. The bar is 2025-06-18 or newer.

  • Pass
    HTTPS

    The endpoint URL is https.

Performance

  • Not measured
    Fast (p50)

    Median latency 796 ms over 1 samples in 7 days. The bar is 1000 ms. Need at least 5 samples.

  • Not measured
    Consistent (p95)

    95th percentile 796 ms over 1 samples in 7 days. The bar is 3000 ms. Need at least 20 samples.

  • Not measured
    Reliable (uptime)

    1 of 1 attempts returned a tool list in 30 days (100%). The bar is 99%. Need at least 30 samples.

  • Not measured
    Stable tool list

    Fewer than two successful probes in 30 days, so tool-list stability is not measured.

Access

  • Pass
    Open access

    tools/list succeeded with no credential, and the record declares none.

  • Not measured
    Standard auth

    Auth is not required, so the OAuth metadata check does not apply.

  • Not measured
    Free tier cited

    No publisher quote has been checked.

Tools

  • Pass
    Tools described

    Every tool description is at least 20 characters.

  • Pass
    Valid input schemas

    Every inputSchema is a JSON Schema object.

  • Pass
    Parameters documented

    73 of 73 parameters have a description (100%). The bar is 80%.

  • Pass
    Safety hints

    Every tool sets readOnlyHint or destructiveHint.

Publisher

  • Pass
    Verified namespace

    Listed in the Official MCP Registry.

  • Fail
    Website matches publisher

    The website host does not match the reverse-domain namespace.

  • Pass
    Source repo linked

    A repository URL is on file.

  • Not measured
    Repo owner matches

    This namespace is not an io.github login, so the owner comparison does not apply.

  • Fail
    Recently updated

    Updated 2026-05-07T04:59:39.740016Z, which is older than 180 days.

  • Fail
    Open-source licence

    GitHub does not report an OSI licence.

Start using Dreamlit

Connect to this server

Choose your client, copy the configuration, then enable the server in your client’s MCP settings.

Run this command in your terminal, then restart Claude Code.

Claude Code

claude mcp add --transport http dreamlit https://mcp.dreamlit.ai/mcp

Capabilities

Tools 16

confirm_publish

Use only after explicit user confirmation and a prior prepare_publish result to publish or schedule a Dreamlit workflow. Side effect: installs live database/repeating/auth triggers, schedules or sends broadcasts, and may enable notification delivery; sandbox mode can hold notifications for inspection. Returns the published workflow status and app URLs. Do not call speculatively or without carrying forward the prepare_publish safety fields.

Discovered
create_or_update_workflow

Creates a new Dreamlit workflow draft or updates an existing draft from an outcome-oriented natural-language prompt. Use after get_status; use get_workflow_and_preview_url first when editing an existing workflow, then describe the steps to change by what they do, not by step id. Each call continues the workflow's current chat; pass startNewChat to start fresh. Broadcasts that are scheduled, sent, or cancelled can't be edited; duplicate them in the Dreamlit web app instead. Existing Supabase Auth workflows can be edited except for the immutable trigger step; creating Supabase Auth workflows must happen through Supabase Auth email setup in the Dreamlit web app. Side effect: may create or modify a draft, but does not publish or install live triggers. Returns the workflow/draft result, action-required or handoff details when more input is needed, and relevant app URLs. Do not use for publishing, direct database changes, or low-level graph edits.

Discovered
edit_workflow_steps

Applies up to 5 exact edits to existing steps in a workflow draft, instantly and without AI: email subject, preview text, HTML, sender, reply-to, and recipient; Slack message; query SQL; wait rule; and step description. Use when you know the exact new value; call get_workflow_and_preview_url first for step ids and current values, with includeEmailHtml to edit HTML. Edits are checked against the draft as it will be once they are all saved: templates and waits may only use fields their step gets, and query edits are rejected if later steps still use fields the new query no longer returns. Each edit applies on its own, except that a SQL edit and the edits below it apply together or not at all. The edits show in the workflow's Dreamlit chat. Side effect: changes the draft only; does not publish. Returns each edit's status with an error code and message for failed edits, and app URLs. Do not use to add, remove, or move steps, edit triggers, or make changes that need writing or judgment; use create_or_update_workflow for those.

Discovered
get_analytics

Read-only. Use get_analytics to query raw notification analytics: email and Slack sends and deliveries, email opens, clicks and bounces, overview metrics, per-recipient rows and workflow run rows, with filters, sorting, and cursor pagination. Returns bounded structured analytics data, effective query metadata, pagination details when rows are included, and relevant app URLs. Use get_lifecycle_report or get_recommendations for Dreamlit's prepared analysis. Does not provide Dreamlit's lifecycle judgments or prioritized recommendations. Do not use for CSV exports, bulk dumps, workflow edits, publishing, or low-level database access.

Discovered
get_lifecycle_report

Side effect: creates a preparation request when no report is ready or refresh is true and the latest report became ready more than 24 hours ago; repeat calls reuse an open request. Pass refresh: true only when the user asks for a new report; when checking on a pending report, call without refresh. Use for Dreamlit's prepared lifecycle analysis: key metrics worth tracking, the signup-to-paid funnel and where users drop off, deliverability health, and which stages (welcome, onboarding, engagement, retention, winback) have no emails yet. Available on Pro and Scale plans only. Reports are prepared by Dreamlit's agents and reviewed by our team. Returns a ready report or a processing state while Dreamlit prepares it; check back later. The requester is also emailed when ready. Does not return raw notification analytics; use get_analytics for that.

Discovered
get_recommendations

Side effect: creates a preparation request when no report is ready or refresh is true and the latest report became ready more than 24 hours ago; repeat calls reuse an open request. Pass refresh: true only when the user asks for a new report; when checking on a pending report, call without refresh. Use for Dreamlit's prioritized recommendations on which emails or workflows to add and what to fix or change in existing ones. Available on Pro and Scale plans only. Reports are prepared by Dreamlit's agents and reviewed by our team. Returns a ready report or a processing state while Dreamlit prepares it; check back later. The requester is also emailed when ready. Does not return raw notification analytics; use get_analytics for that.

Discovered
get_status

Read-only. Use first when the agent needs Dreamlit product guidance, prompting guidance, approved workspace context, project setup, schema hints, workflow state, or relevant app URLs. Returns a compact context pack with concepts, recommended tool flow, actor/workspace/project data, optional authoring context, optional workflow context, and appUrls. Workflow summaries include liveStatus, and broadcasts include an authoritative broadcast.state. Database tables that Dreamlit can see but not read rows from (row level security with no policy for Dreamlit) carry a warning and fixSql; run fixSql, or ask the user to run it, in the Supabase SQL editor. If Dreamlit couldn't read a connection's tables, that connection has tablesError instead of an empty list meaning no tables. Do not use this to create, update, publish, or unpublish workflows.

Discovered
get_workflow_and_preview_url

Read-only. Use before editing, verifying, or preparing publish for an existing workflow. Returns normalized draft and published workflow structure, liveStatus, the Dreamlit preview/builder URL, and status URLs. Broadcasts also include a broadcast object with state (draft, scheduled, sending, sent, cancelled, or failed), the run's scheduledFor, and cancelledAt or completedAt; this, not published or the trigger's sendAt, says whether the broadcast will send. Email HTML bodies are left out unless includeEmailHtml is true. Pass step ids only to edit_workflow_steps; when asking create_or_update_workflow for changes, describe steps by what they do. Do not use to create, update, publish, or unpublish workflows.

Discovered
list_brand_styles

Read-only. Use when the user wants to inspect saved project brand styles for generated email drafts. Returns paginated project brand style summaries, ids, default status, style metadata, and app URLs. Do not use to create, update, delete, or expose raw style provenance.

Discovered
list_projects

Read-only. Use when the target project is unknown or the user names a project instead of providing an id. Returns accessible Dreamlit projects in the approved workspace with project and workflow-list URLs. Do not use to inspect workflow structure or author changes.

Discovered
list_workflows

Read-only. Use to find workflows in a project by name, description, or trigger type before inspection or editing. Trigger filters include database, auth email, repeating, broadcast, and no-trigger workflows. Returns paginated workflow summaries, published/sandbox state, liveStatus, trigger type, folder (the folder name in the list the workflow shows in, or null), workflow URLs, totalCount, hasMore, and nextOffset. Broadcasts also include a broadcast object with state (draft, scheduled, sending, sent, cancelled, or failed), scheduledFor, and cancelledAt or completedAt; for broadcasts, broadcast.state and liveStatus are authoritative over published, because a cancelled broadcast stays published but will not send. Do not use as the final source of truth before editing; call get_workflow_and_preview_url for full structure.

Discovered
move_workflow_to_folder

Use when the user asks to put a workflow or broadcast in a folder, move it to another folder, or take it out of a folder. Folders are per list: the Published and Drafts lists on the Workflows page and the Sent and Drafts lists on the Broadcasts page each have their own folders, and this moves the workflow within the list it shows in now. Folders are flat; a "/" is just part of the name. A draft's folder does not carry over when it is published. Side effect: may create the folder; never deletes folders and does not change the draft, published version, or live triggers. Returns the folder, the list, and app URLs. Do not use to rename or delete folders.

Discovered
prepare_publish

Read-only validation step before publishing. Use after the user asks to publish and after inspecting the workflow if needed. Returns validation status, publish mode, preview URL, and a confirmation payload with version/timestamp safety checks for confirm_publish. Does not publish or schedule anything by itself.

Discovered
rename_workflow

Use when the user asks to rename an existing Dreamlit workflow. Side effect: changes only the workflow name; the draft, published version, and live triggers are unchanged. Returns the updated name and app URLs. Do not use to edit workflow content or to name a new workflow; pass name to create_or_update_workflow instead.

Discovered
send_workflow_test

Side effect when confirmSend is true: sends one current draft email or Slack message for review, but does not publish, schedule, or edit the workflow. Use after authoring or inspection when the user wants to verify an important message before publishing. Returns sent-test details, app URLs, a confirmation request, or a clarification request with user-facing message choices. If the test has no data because row level security hides a source table's rows from Dreamlit, returns error code table_rows_hidden_by_rls with unreadableTables[].fixSql to run, or ask the user to run, in the Supabase SQL editor. Do not expose test target ids as copy; ask users to choose by message label and confirm before sending.

Discovered
unpublish_workflow

Use after explicit user intent to unpublish a Dreamlit workflow. Side effect: disables live triggers or schedules so no new runs start. Runs already in progress keep sending unless inProgressRuns is "stop". If the workflow has in-progress runs and inProgressRuns is omitted, returns clarification_required with the count and changes nothing: ask the user whether to stop those runs or let them finish, then call again with their choice. Returns updated workflow status, stopped run counts (allRunsStopped, remainingRunCount), and app URLs. If some runs could not be stopped, ask the user, then call again with inProgressRuns "stop": on an already-unpublished workflow it only stops the runs still in progress. Do not use for deleting drafts, canceling one broadcast run, or editing workflow content.

Discovered

Recent measurements

Probe history

View status
Latest stored probes. Raw probe rows are kept for 14 days.
CheckedResultResponseTools
Connected796 ms16