@thalovant/mcp-server 0.3.0 → 0.4.0

package/CHANGELOG.md

@@ -1,5 +1,11 @@
 # Changelog
 
+## 0.4.0 — 2026-09-12
+
+- Add locale-aware sentence inventory examples with SDK 0.6.0, canonical slot defaults and unique rendered limits.
+- Handle SIGTERM/SIGINT explicitly and close MCP sessions and HTTP connections when running as PID 1.
+- Align the Dockerfile version label with the package and registry metadata.
+
 ## 0.3.0 — 2026-09-12
 
 - Use Node SDK 0.5.0 request hints, bounded embedded audio replies, and speakable intent examples. Keep embedded audio out of model text; offer explicit audio/resource content blocks through includeAudio.

package/README.md

@@ -384,7 +384,7 @@ Destructive, **not registered unless explicitly enabled** (see [Destructive Tool
 Tool outputs redact credential-shaped fields. `thalovant_create_client_identity` does not return secret identity material; pass `savePath` when you want the full identity written to a local file with mode `0600`. `savePath` is confined to the server's identity directory (`THALOVANT_MCP_IDENTITY_DIR`, default `<config-dir>/thalovant/identities`): pass a plain filename, since absolute paths outside that directory and `..` traversal are rejected, so a model cannot drop a credential file into a git working tree or synced folder. `thalovant_config_status` reports the active `identityDir`.
 
 `thalovant_intent_inventory` uses the runtime identity and accepts `languages`,
-`describe`, `fallback`, `speakable`, `slots`, `exampleLimit`, and a per-query/batch `timeoutMs`. It returns registered
+`describe`, `fallback`, `speakable`, `sentence`, `slots`, `exampleLimit`, and a per-query/batch `timeoutMs`. It returns registered
 intents and examples, fallback skills, `fallbacks_known`, and `may_answer` by
 requested language. A missing or denied optional fallback-skill query remains
 unknown, rather than being reported as a known empty list. The optional probe
@@ -464,7 +464,7 @@ Runtime calls sharing the same hub client identity run sequentially within one
 MCP process. Use a distinct client identity for each independently running MCP
 server so the hub can keep their sessions separate.
 
-Version 0.3.0 requires `@thalovant/sdk` `^0.5.1` (0.5.1 through versions below 0.6.0). This release enforces
+Version 0.4.0 requires `@thalovant/sdk` `^0.6.0` (0.6.0 through versions below 0.7.0). This release enforces
 secure effective MQTT URLs and carries a single connection deadline through
 MQTT setup and HTTP failure cleanup. Runtime tools support
 HiveMind v3 Noise over WSS, HTTPS and MQTT over TLS. `thalovant_healthcheck`
@@ -557,3 +557,14 @@ never repeats an accepted mutation and starts no new read after its deadline;
 an already-running HTTP request retains its normal request timeout.
 
 Read history with `thalovant_list_hub_skill_history` (`hubId`, optional `limit`). This tool remains available in read-only mode.
+
+
+Inventory `sentence: true` implies speakable rendering and capitalizes/punctuates
+examples using bundled thalovant-languages 0.1.1 rules. Regional locale matching
+follows OVOS distances; explicit `slots` override locale sample values. Complete
+phrases rank first, and `exampleLimit` counts unique nonempty rendered results.
+Raw intent definitions remain in the response; unknown locales retain bare text.
+
+The CLI handles SIGTERM and SIGINT in stdio and HTTP modes, including when it
+runs as the container's PID 1. It closes sessions and HTTP connections before a
+successful exit, with a ten-second failure deadline for stalled cleanup.

package/dist/index.js

@@ -1346,7 +1346,8 @@ export function createServer() {
         inputSchema: {
             ...runtimeAuthSchema,
             languages: z.array(z.string().trim().min(1).max(100)).min(1).max(20).default(["en-us"]),
-            speakable: z.boolean().default(false).describe("Render intent patterns into example sentences; preserve whole-sentence priority."),
+            speakable: z.boolean().default(false).describe("Render intent patterns with locale slot examples; preserve complete-phrase priority."),
+            sentence: z.boolean().default(false).describe("Capitalize and punctuate examples using the selected locale. Implies speakable rendering."),
             slots: z.record(z.string().max(500)).optional().describe("Example values for named intent slots."),
             exampleLimit: z.number().int().min(1).max(20).default(2),
             describe: z.boolean().default(true),
@@ -1354,14 +1355,14 @@ export function createServer() {
             timeoutMs: z.number().int().min(1_000).max(MAX_TIMEOUT_MS).default(DEFAULT_TIMEOUT_MS).describe("Per-query/batch timeout; the optional fallback-skill probe is capped at 1500ms."),
         },
         annotations: { readOnlyHint: true, openWorldHint: true },
-    }, async ({ languages, describe, fallback, timeoutMs, speakable, slots, exampleLimit, ...runtime }) => withRuntimeClient(runtime, async (client, signal) => {
+    }, async ({ languages, describe, fallback, timeoutMs, speakable, sentence, slots, exampleLimit, ...runtime }) => withRuntimeClient(runtime, async (client, signal) => {
         await client.connect(undefined, signal);
         throwIfRuntimeCancelled(signal);
         const inventory = await client.intents(languages, { describe, fallback, timeoutMs: clampTimeout(timeoutMs) });
         return jsonContent(redactSecrets({
             ...inventory.asObject(),
             examples: inventory.intents.map(intent => ({ id: intent.id,
-                languages: Object.fromEntries(inventory.languages.map(lang => [lang, intent.examples(lang, exampleLimit, { speakable, slots })])),
+                languages: Object.fromEntries(inventory.languages.map(lang => [lang, intent.examples(lang, exampleLimit, { speakable, sentence, slots })])),
             })),
             may_answer: Object.fromEntries(inventory.languages.map(lang => [lang, inventory.mayAnswer(lang)])),
         }));
@@ -2322,8 +2323,13 @@ export async function startHttpServer(config = getHttpConfig()) {
         }
     }, Math.min(config.sessionTtlMs, 60_000));
     cleanupTimer.unref();
+    let closing;
     const httpServer = createHttpServer(async (req, res) => {
         setSecurityHeaders(res);
+        if (closing) {
+            sendJson(res, 503, { error: "Server is shutting down." });
+            return;
+        }
         try {
             if (!validateHost(req, config)) {
                 sendJson(res, 403, { error: "Forbidden host." });
@@ -2366,11 +2372,19 @@ export async function startHttpServer(config = getHttpConfig()) {
                 return;
             }
             const auth = await authenticate(req, config);
+            if (closing) {
+                sendJson(res, 503, { error: "Server is shutting down." });
+                return;
+            }
             const authenticatedReq = req;
             authenticatedReq.auth = authInfoFromAuthResult(auth);
             authenticatedReq.auth.resource = config.resourceUrl;
             if (req.method === "POST") {
                 const body = await readJsonBody(req, config.maxBodyBytes);
+                if (closing) {
+                    sendJson(res, 503, { error: "Server is shutting down." });
+                    return;
+                }
                 const sessionId = typeof req.headers["mcp-session-id"] === "string" ? req.headers["mcp-session-id"] : undefined;
                 if (sessionId) {
                     const session = sessions.get(sessionId);
@@ -2425,6 +2439,11 @@ export async function startHttpServer(config = getHttpConfig()) {
                     });
                 };
                 await mcpServer.connect(transport);
+                if (closing) {
+                    await mcpServer.close();
+                    sendJson(res, 503, { error: "Server is shutting down." });
+                    return;
+                }
                 await transport.handleRequest(authenticatedReq, res, body);
                 return;
             }
@@ -2483,12 +2502,12 @@ export async function startHttpServer(config = getHttpConfig()) {
     console.error(`Thalovant MCP Streamable HTTP listening at ${url}`);
     return {
         url,
-        close: async () => {
+        close: () => {
+            if (closing)
+                return closing;
             clearInterval(cleanupTimer);
-            for (const sessionId of Array.from(sessions.keys())) {
-                await closeSession(sessionId);
-            }
-            await new Promise((resolveClose, rejectClose) => {
+            // Stop admission before closing sessions; then release remaining HTTP sockets.
+            const stopped = new Promise((resolveClose, rejectClose) => {
                 httpServer.close((error) => {
                     if (error)
                         rejectClose(error);
@@ -2496,9 +2515,44 @@ export async function startHttpServer(config = getHttpConfig()) {
                         resolveClose();
                 });
             });
+            closing = Promise.all([
+                stopped,
+                (async () => {
+                    try {
+                        await Promise.all(Array.from(sessions.keys(), closeSession));
+                    }
+                    finally {
+                        httpServer.closeAllConnections();
+                    }
+                })(),
+            ]).then(() => undefined);
+            return closing;
         },
     };
 }
+// CLI signal handling is explicit because Node may run as PID 1 in the OCI image.
+function handleShutdown(close) {
+    let stopping = false;
+    const shutdown = () => {
+        if (stopping)
+            return;
+        stopping = true;
+        const deadline = setTimeout(() => {
+            console.error("MCP shutdown exceeded 10 seconds.");
+            process.exit(1);
+        }, 10_000);
+        void Promise.resolve().then(close).then(() => {
+            clearTimeout(deadline);
+            process.exit(0);
+        }, (error) => {
+            clearTimeout(deadline);
+            console.error("Error during MCP shutdown:", error);
+            process.exit(1);
+        });
+    };
+    process.on("SIGTERM", shutdown);
+    process.on("SIGINT", shutdown);
+}
 export async function main() {
     const requestedTransport = process.argv.includes("--http")
         ? "http"
@@ -2506,12 +2560,14 @@ export async function main() {
             ? "stdio"
             : (process.env.MCP_TRANSPORT ?? "stdio");
     if (requestedTransport === "http" || requestedTransport === "streamable-http") {
-        await startHttpServer();
+        const http = await startHttpServer();
+        handleShutdown(http.close);
         return;
     }
     const transport = new StdioServerTransport();
     const server = createServer();
     await server.connect(transport);
+    handleShutdown(() => server.close());
 }
 function isMainModule() {
     if (!process.argv[1])

package/examples/intent-listing.json

@@ -0,0 +1,15 @@
+{
+  "name": "thalovant_intent_inventory",
+  "arguments": {
+    "identityFile": "/path/to/identity.json",
+    "languages": [
+      "fr-CA",
+      "en-US"
+    ],
+    "sentence": true,
+    "slots": {
+      "location": "Montréal"
+    },
+    "exampleLimit": 3
+  }
+}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@thalovant/mcp-server",
-  "version": "0.3.0",
+  "version": "0.4.0",
   "description": "Stdio and Streamable HTTP MCP server for Thalovant control-plane and hub runtime APIs.",
   "license": "MIT",
   "type": "module",
@@ -64,7 +64,7 @@
   },
   "dependencies": {
     "@modelcontextprotocol/sdk": "^1.30.0",
-    "@thalovant/sdk": "^0.5.1",
+    "@thalovant/sdk": "^0.6.0",
     "jose": "^6.2.8",
     "zod": "^3.25.76"
   },

package/server.json

@@ -7,19 +7,19 @@
     "url": "https://github.com/thalovant/thalovant-mcp",
     "source": "github"
   },
-  "version": "0.3.0",
+  "version": "0.4.0",
   "packages": [
     {
       "registryType": "npm",
       "identifier": "@thalovant/mcp-server",
-      "version": "0.3.0",
+      "version": "0.4.0",
       "transport": {
         "type": "stdio"
       }
     },
     {
       "registryType": "oci",
-      "identifier": "ghcr.io/thalovant/thalovant-mcp:0.3.0",
+      "identifier": "ghcr.io/thalovant/thalovant-mcp:0.4.0",
       "transport": {
         "type": "streamable-http",
         "url": "http://127.0.0.1:3000/mcp"