appPackage manifestv3.0.5

app.communicationos-mcp

MCP server exposing CommunicationOS contracts as AI-callable tools via the Model Context Protocol.

@lssm-tech/app.communicationos-mcp
Source, release and verification context

Source snapshot bcc79fcc · Declared package version 3.0.5. This page combines the checked-in package guide and manifest; commands and runtime behavior were not executed in this review.

Package source ↗ Manifest ↗
servernode
EXPLORE THE PACKAGE

See where this package fits.

Follow a declared connection, inspect an import, then open the source guide when you need the detail.

THIS PACKAGEapp.communicationos-mcpapps

Arrows point from a package to its direct dependency. Each count is derived from the captured manifests.

11 déclarations · les 6 premières affichéesSELECT TO INSPECT

CommunicationOS MCP

Governed Model Context Protocol surface for CommunicationOS. It exposes tenant-scoped thread reads, drafting, summarization, extraction, provider health, inbound-route inspection, and the canonical Forms operation registry. Forms tools inherit their contract auth and PII annotations. Direct communication.reply.send is intentionally excluded: outbound delivery remains subject to room membership, current authority, and the approval/send orchestrator.

MailerSend diagnostics

createMailerSendMcpProviderDependencies binds the read-only integrations.email.provider.health and integrations.email.inbound-routes.list tools. The host injects a tenant-and-connection-scoped resolver returning MailerSend credentials and a domain id. Secrets never appear in MCP tool inputs or outputs.

Both tools take tenantId as input, but they only run for the verified subject (src/provider-scope.ts). The input tenant must equal the subject's tenantId, and the subject must hold the admin role, because the contracts declare policy.auth: 'admin' and the MCP gate itself only checks for an authenticated actor. Anything else is FORBIDDEN (403) before the connection resolver is called, so a token for one tenant can never load another tenant's MailerSend connection.

Subject and trust boundary

The stdio entrypoint (startCommunicationOsMcpServer) verifies one CommunicationOS access token at startup with createCommunicationAccessTokenVerifier from @lssm-tech/integration.communication-os-host-auth, then builds the server. It reads only these variables:

Variable Objectif
COMMOS_MCP_ACCESS_TOKEN Access token minted by communicationos-api (issuer https://api.hcircle.app, audience communication-os-api).
COMMOS_MCP_TENANT_ID, COMMOS_MCP_WORKSPACE_ID, COMMOS_MCP_SPACE_ID The scope the token must carry exactly.
COMMUNICATION_OS_ACCESS_TOKEN_SECRET The shared HS256 signing secret (at least 32 bytes).

A verified, in-scope token becomes the subject of every tool call: { actor: 'user', userId: <principalId>, tenantId, workspaceId, roles }, with the token's space as the policyContext.spaceId fact. Handlers and policies therefore see the verified workspace and space, not only the tenant. The principal and the token's exp are cached; from exp onward, and whenever a variable is missing or the secret, signature, issuer, audience, scope or expiry does not verify, every call is anonymous and the MCP auth gate rejects it. The startup log line reports subject: authenticated | unconfigured | rejected and never the token. Restart the gateway with a fresh token to renew it.

Trust boundary. Verifying the token needs the shared HS256 secret, which can also mint tokens. Run this gateway only server-side, next to communicationos-api and inside its secret boundary, never on a client or developer machine. Distributing it to clients would need a contract-first token introspection route instead of the shared secret.

The import-time server export and buildCommunicationOsMcpServer() without a resolver stay anonymous. startCommunicationOsMcpServer builds and connects its own gateway after verification, so the exported server is not the connected instance. The default server binds no provider resolver, so production hosts must also inject a vault-backed connection resolver before the provider tools can execute; the subject guard above still applies.

Declared capabilities.

This manifest does not declare a capability catalogue. Use the guide, import map and source to assess the package.

Imports and export paths.

4 entries

These are package-manifest entry points. Named symbols listed above come from the capability catalogue; this is not an exhaustive TypeScript symbol reference.

4 imports

Dependencies and related packages.

Runtime dependencies 11
@modelcontextprotocol/sdk
catalog:ai
Development dependencies 3
@types/node
catalog:types
typescript
catalog:

Package scripts.

Commands declared by this package. Review its setup requirements before running them; no command here was executed by this review.

startbun src/index.ts
typechecktsgo --noEmit
lint:checknode ../../../scripts/biome.cjs check .
lintbun run lint:fix
lint:fixnode ../../../scripts/biome.cjs check --write .
Votre vue

Make room for your context.

Shape the presentation around this moment. You can change it whenever you need.

Votre contexte

The different parts of you

Choose a starting point

Un mélange utile de contexte, de visuels et de guidage discret.

Tune each dimension
Each dimension can change independently. Product permissions and available actions stay unchanged.

Support the way you begin and return

Tune your support

Confort de lecture

Choisir votre langue

La langue de l’interface change uniquement le texte. Les marchés, les prix et les parcours conservent leurs propres paramètres.

Couverture des traductions

Une partie du contenu est affichée en anglais.

Navigation151
Graphe616
Catégories378
Portfolio833
Exemples2 035
Aperçu de la source1 542

D’autres traductions seront ajoutées à mesure que le contenu sera relu.

Cette préférence est enregistrée dans ce navigateur.

Rechercher sur le site

Des points de départ utiles

Aucun résultat. Essayez une expression plus courte, un nom d’opération ou un autre périmètre de recherche.

UN PEU DE PLACE POUR EXPLORER

Trouvez quelque chose à explorer.

Ouvrez un aperçu détaillé sur cette page, essayez une autre perspective ou suivez le détail vers sa propre page.

Fonctionnement des aperçus et des commandes
01

Regarder avant de partir

Survolez un lien ou sélectionnez-le avec Tab pour voir un aperçu. Cliquer sur le lien d’origine ou appuyer sur Entrée ouvre toujours sa destination.

02

Rester dans le contexte

Entrez dans l’aperçu pour explorer ses sections. Depuis un lien sélectionné, utilisez Alt + ↓ pour y entrer au clavier. Échap le ferme.

03

Ouvrir les possibilités

Les lignes dépliables, les choix de perspective et les commandes mises en évidence révèlent davantage. Votre vue adapte le confort de lecture et le contexte exploré.

Un bouton séparé pour voir de plus près

Affichez de petits boutons d’aperçu à côté des liens quand l’espace le permet. Touchez l’œil pour examiner ou le lien pour y aller directement. Ce choix reste dans cet onglet.

Aucun historique de navigation n’est collecté. La réduction des animations suit votre appareil et Votre vue.