1. Conduid
  2. Healthcare
  3. dooor-os-mcp
MCP server · Healthcare

dooor-os-mcp

Exposes Dooor OS workspace capabilities to MCP-compatible AI clients, enabling natural language querying of data sources, app and deployment inspection, agent and database management, and governance trace reading.

Unclaimed healthcare
34Low

Scored 3 months ago · breakdown

About dooor-os-mcp

dooor-os-mcp is an MCP server in the Healthcare category: exposes Dooor OS workspace capabilities to MCP-compatible AI clients, enabling natural language querying of data sources, app and deployment inspection, agent and database management, and governance trace reading.

Install

From source
git clone https://github.com/Dooor-AI/dooor-os-mcp

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Tools (85)

attach_databaseAttach a managed database to an app. SQLITE: mounts the PVC at the given path (default /var/lib/dooor/sqlite) and injects DATABASE_URL=file:<path>/db.sqlite via envFrom Secret on the next deploy; the app is pinned to maxReplicas=1. POSTGRES/REDIS: records the link; auto-injection of the connection string into env is on the roadmap.

Parameters not published.

capabilitiesWhoami plus a compact map of this MCP server: active workspace, API key scopes, tool families, and optional read-only probes for connected data sources. Use first when you need to know what this key can access.

Parameters not published.

chat_with_agentSend a message to an agent and get a response. Pass sessionId to continue a conversation.

Parameters not published.

create_agentCreate a new AI agent. Use a template for quick setup, or configure from scratch.

Parameters not published.

create_api_keyCreate a new API key. Scopes are mandatory and should be kept minimal. The secret is returned only once.

Parameters not published.

create_appCreate a new app in the workspace. Connect a Git repo and configure build settings.

Parameters not published.

create_databaseProvision a new managed database. POSTGRES = full Cloud Native PG cluster (separate Pod, network-accessible). SQLITE = file-based, mounted into the app pod via PVC (single-writer, RWO; the consuming app must be pinned to maxReplicas=1). REDIS = key-value store via StatefulSet. For demos and single-pod apps without scale-out needs, prefer SQLITE because it provisions instantly without Longhorn HA pressure.

Parameters not published.

data_askPRIMARY data tool. Ask a natural-language business question (PT-BR or EN) about the workspace

Parameters not published.

data_connection_capabilitiesDiscover the allowlisted read-only entities, list/get operations, typed fields, pagination contract and fixed filter keys for one live connection. Fixed filters are enforced by Dooor and cannot be overridden by the caller.

Parameters not published.

data_connection_readExecute one allowlisted read-only list/get operation through Dooor. Call data_connection_capabilities first and use exactly one entity and operation it returns. This tool never exposes source credentials and never performs source writes.

Parameters not published.

Show all 85 tools
data_connectionsList the live operational data connections visible to this key. Returns neutral connection IDs, names, types and status without credentials. Use the selected sourceId with data_connection_capabilities before reading it.

Parameters not published.

data_insightsGet the latest proactive insight digest discovered across the connected sources (ranked findings with metric, value, evidence and recommended action). Read-only.

Parameters not published.

data_overviewGet the aggregated overview exposed by the active data product for this workspace. The available domains and metrics are product-defined. Call data_products first when the workspace is new. Read-only.

Parameters not published.

data_productsList the productized data experiences enabled in this workspace, including capabilities, REST paths, MCP tools and backing stores. Use this before choosing data_* or lake_* tools for a new workspace.

Parameters not published.

data_sourcesList the governed sources exposed by the active data product, including the metadata and statistics its provider makes available. The key field is the only supported relation name for data_sql. For replicated-source freshness, lastSyncAt is authoritative when present. If the provider exposes no field explicitly labeled as freshness, report freshness as unknown and do not claim a sync ran or failed to run. Read-only.

Parameters not published.

data_sqlRun one ad-hoc read-only SQL query over the workspace-scoped business relations exposed by the active data product. Inspect data_products and data_sources first, then use only the key field returned by data_sources as each relation name. Do not infer sync freshness from business dates or row createdAt/updatedAt values; use data_sources.lastSyncAt or other provider-labeled freshness metadata. Platform tables, physical table names and mutating statements are blocked; execution and row limits are enforced server-side. Read-only.

Parameters not published.

data_tablePreview rows from one source exposed by the active data product provider. Call data_products first and use this tool only when data_table is advertised by an enabled capability. Row timestamps and business dates are not evidence of sync freshness; use explicit freshness metadata from data_sources. Read-only.

Parameters not published.

delete_appDelete an app. Use permanent=true to permanently remove all data.

Parameters not published.

delete_databasePermanently delete a managed database and all its data. This cannot be undone.

Parameters not published.

delete_env_varDelete an environment variable from an app.

Parameters not published.

deploy_agentTrigger an agent deploy. Returns immediately with a deployment record (status RECONCILING). The pod readiness check runs asynchronously - DO NOT report success or failure based on this response alone. To know whether the deploy actually succeeded, poll get_agent_deployment until status reaches READY/FAILED/TIMEOUT.

Parameters not published.

deploy_appTrigger a new deployment for an app using the App

Parameters not published.

deploy_app_from_directoryTar a local directory (respecting .dockerignore/.gitignore + safe defaults), upload it, and trigger a deploy. Use this when the user wants to ship an app from their local filesystem without pushing to git. The MCP runs locally so it has filesystem access. Returns the new deployment id.

Parameters not published.

deploy_app_from_imageTrigger a deploy from a pre-built container image (no build step). Use when the user already published an image to a registry. The image must be pullable by the cluster; for private registries pass registryAuthRef pointing to credentials in Vault.

Parameters not published.

deploy_app_from_tarballUpload an already-built .tar.gz file and trigger a deploy. Use when the user has a prepared tarball; otherwise prefer deploy_app_from_directory which handles tar creation.

Parameters not published.

deploy_app_upload_finalizeStep 2 of 2 of the hosted upload deploy: after the tarball was PUT to the presigned URL from deploy_app_upload_init, verifies the upload (SHA-256 must match) and triggers the deployment from it. Returns the created deployment; poll get_deployment until buildStatus and deployStatus reach ACTIVE or FAILED.

Parameters not published.

deploy_app_upload_initStep 1 of 2 to deploy a LOCAL project directory when deploy_app_from_directory is NOT in this server

Parameters not published.

detach_databaseDetach a database from an app. The database itself stays provisioned and its data is preserved; only the link is removed. The next deploy will not mount the volume or inject the connection Secret.

Parameters not published.

get_agentGet detailed information about a specific agent.

Parameters not published.

get_agent_deploymentGet a specific agent deployment, including K8s pod events captured on failure. Use this AFTER deploy_agent to know whether the deploy actually succeeded - the response includes status (PENDING/RECONCILING/READY/FAILED/TIMEOUT), errorReason, errorMessage, and (when failed) k8sEvents with the actual pod-level cause (ImagePullBackOff, CrashLoopBackOff, etc).

Parameters not published.

get_appGet detailed information about a specific app, including deployment status and config.

Parameters not published.

get_app_healthGet health status of an app (healthy, degraded, unhealthy, unknown).

Parameters not published.

get_app_logsGet application logs. Useful for debugging runtime issues.

Parameters not published.

get_app_metricsGet app performance metrics (CPU, memory, requests, latency, errors).

Parameters not published.

get_app_platform_domainGet the app

Parameters not published.

get_app_sourceGet the App source configuration (GIT repo / UPLOAD slot / IMAGE reference).

Parameters not published.

get_app_statsGet summary statistics for all apps in the workspace (total, active, failed, etc.).

Parameters not published.

get_build_logsGet build logs for a deployment. Useful for debugging failed builds.

Parameters not published.

get_databaseGet details of a managed database.

Parameters not published.

get_database_connectionGet connection credentials for a database (URI, username, password, host, port).

Parameters not published.

get_database_statusGet runtime status of a database (ready/desired pods).

Parameters not published.

get_deploymentGet details of a specific deployment.

Parameters not published.

get_git_install_urlGet the GitHub App installation URL. Redirect users here to connect their GitHub org.

Parameters not published.

get_pipeline_stateGet the full pipeline state for an app (source, build, deploy, harbor, runtime status).

Parameters not published.

get_runtime_statusGet live Kubernetes runtime status for a deployment (pods, phase, conditions).

Parameters not published.

get_workspace_overviewGet a high-level overview of the workspace: total apps, deploys today, health summary, costs.

Parameters not published.

harbor_observabilityRead the AI governance data (Harbor) for an app: traces of every AI call,

Parameters not published.

integration_guideHow to consume this workspace

Parameters not published.

lake_askAsk a natural-language question (PT-BR or EN) over the active product

Parameters not published.

lake_browseBrowse raw or curated rows from a layer exposed by the active product provider. Returns typed columns, rows, executed SQL and server-side query time. Discover valid identifiers with lake_sources first. Read-only, max 1000 rows.

Parameters not published.

lake_catalogGet the active product

Parameters not published.

lake_code_listBrowse the indexed code chunks page by page (no query). Returns { items, total, nextOffset }. Pass the opaque nextOffset cursor from a previous call to page forward. Use lake_code_search instead when you know what you are looking for. Read-only.

Parameters not published.

lake_code_searchSemantic search over the business-rule source code indexed for the active product. Returns relevant code chunks with file path, summary, snippet and score. Read-only.

Parameters not published.

lake_dashboardGenerate a multi-panel analytical dashboard from a natural-language brief. The active product provider selects governed metrics and dimensions from its catalog, executes the panels, and returns data ready to render. Read-only.

Parameters not published.

lake_queryRun a structured, validated aggregation over the active product

Parameters not published.

lake_sourcesPage or summarize the raw and curated layers exposed by the active product

Parameters not published.

lake_sqlRun one ad-hoc read-only SQL statement over the active product

Parameters not published.

list_agent_deploymentsList recent deploy attempts for an agent (most recent first). Use to find a deploymentId to inspect or to see deploy history.

Parameters not published.

list_agent_templatesList available agent templates for quick agent creation.

Parameters not published.

list_agentsList AI agents in the workspace.

Parameters not published.

list_alertsList monitoring alerts for the workspace or a specific app.

Parameters not published.

list_api_keysList all API keys in the workspace (secrets are never returned).

Parameters not published.

list_app_databasesList databases attached to an app, including the mount path (for SQLite) and the connection Secret name.

Parameters not published.

list_appsList all apps in the workspace. Optionally filter by status, type, or search term.

Parameters not published.

list_branchesList branches for a specific repository.

Parameters not published.

list_databasesList managed databases in the workspace.

Parameters not published.

list_deploymentsList deployment history for an app.

Parameters not published.

list_env_varsList all environment variables for an app. Secret values are masked.

Parameters not published.

list_git_installationsList connected GitHub installations for the workspace.

Parameters not published.

list_reposList repositories available from a GitHub installation.

Parameters not published.

list_revisionsList all revisions for an app with current traffic allocation percentages.

Parameters not published.

query_databaseRun a read-only SQL query against a managed POSTGRES database and get the rows back. Only SELECT, WITH, SHOW, DESCRIBE and EXPLAIN are allowed; writes and DDL are rejected server-side. A LIMIT is injected when absent and results are row-capped. Every call is audited. Requires the databases:query scope and the database to be RUNNING.

Parameters not published.

remove_app_platform_domainRelease the app

Parameters not published.

restart_agentTrigger a rolling restart of an agent container.

Parameters not published.

revoke_api_keyPermanently revoke an API key. Cannot be undone.

Parameters not published.

rollback_deployRollback to a specific deployment. Creates a new deployment from the specified version.

Parameters not published.

scale_appScale an app to a specific number of replicas. Use replicas=0 to stop the app.

Parameters not published.

set_app_platform_domainReserve and route a globally unique managed alias such as finance-copilot.apps.dooor.ai. No customer DNS change is required. This mutates platform routing and requires apps:write.

Parameters not published.

set_app_sourceConfigure or replace the App

Parameters not published.

set_env_varsBulk set environment variables for an app. Existing keys are updated, new keys are created.

Parameters not published.

set_trafficSet traffic allocation across revisions. Percentages must sum to 100. Cloud Run supports weighted splits; the current K8s runtime only supports a single 100% active revision.

Parameters not published.

stop_agentStop a running agent (scales container to 0).

Parameters not published.

sync_env_varsSync environment variables to the running containers without a full redeploy.

Parameters not published.

update_agentUpdate an agent

Parameters not published.

update_appUpdate an app

Parameters not published.

Ask AI

✦

Ask AI about dooor-os-mcp

Powered by Claude · Grounded in docs

I know everything about dooor-os-mcp. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

6 checks have not been run yet.

Questions

About dooor-os-mcp

How do I install dooor-os-mcp?

No package has been published for this server, so install it from source: clone the repository and follow its README.

Is dooor-os-mcp safe to use with an AI agent?

Its trust score is 34 out of 100 (low). Conduid hasn't run static security checks on this repository yet, so review the source yourself before granting it credentials. It has no ConduID identity yet, so agent calls to it are not receipted.

Is dooor-os-mcp still maintained?

Conduid hasn't recorded a commit date for this repository yet. Check the repository directly for recent activity.