@talqing/mcp 0.14.0 → 0.16.0

package/SKILL.md

@@ -89,9 +89,10 @@ An agent is a single object, `config`, plus what you attach to it:
   the request that starts a session supplies values that override the declared
   defaults. See **`vars`** below.
 - **`llm`, `stt`, `tts`** — `{provider, model}` from the provider catalog. Text
-  agents have only `llm`; voice and video need all three.
+  agents have `llm`, plus an optional `stt` that transcribes voice notes (see
+  **Images and files**); voice and video need all three.
 - **`language`** — the one language the caller and the agent speak, shared by
-  every model. `null` is Auto. Voice and video only.
+  every model. `null` is Auto. On a text agent, only with `stt` set.
 - **`timezone`** — an IANA name such as `Asia/Kolkata`. What the three
   `{{system_vars.…}}` clock variables resolve against, on every channel. `null`
   until one of them is used, and then required. Set it whenever the agent has to
@@ -362,8 +363,9 @@ support the channel, a language the model does not offer, a speed outside the
 model's range, an unpublished tool — all are refused with an explanation. Read
 the error and correct it; that is faster than guessing.
 
-Some fields are normalized for you. Switching to `text` clears `stt`, `tts`,
-`greeting`, `turn_handling` and `avatar`. Switching to `video` fills in an
+Some fields are normalized for you. Switching to `text` clears `tts`,
+`greeting`, `turn_handling` and `avatar`, and keeps `stt` (voice notes on) —
+set it to `null` in the same write unless the user wants voice notes. Switching to `video` fills in an
 `avatar`, and the call ignores `resume_false_interruption` — the avatar cannot
 resume a sentence it has already stopped rendering. Keypad input is voice-only
 and a text agent cannot watch a screen share, so enabling either off its channel
@@ -380,9 +382,14 @@ voice. Read it before choosing; do not assume a model exists.
 
 People can send files on WhatsApp (Twilio senders) and in a web chat, and a photo
 on a web voice or video call. Every file is stored on the conversation; photos,
-PDFs and Word (`.docx`) files are also read by the agent, and anything else
-(voice notes, videos, spreadsheets) is named to it as something it cannot open.
-**There is no setting for this.** Whether the agent sees a photo or a scanned
+PDFs and Word (`.docx`) files are also read by the agent, a shared contact card
+is read as its name, numbers and emails, and anything else (videos,
+spreadsheets) is named to it as something it cannot open.
+**Voice notes** are heard only by a text agent with `stt` set (off by default):
+each note is transcribed in one request on that model, so only entries with
+`file_transcription` set qualify. Turn it on when the user's customers send voice
+notes — on WhatsApp in India, nearly always. A call never transcribes one.
+**There is no setting for images.** Whether the agent sees a photo or a scanned
 PDF page is decided by the model it runs: `vision` on its catalog entry. Most
 entries read images; `gemini-3.8-flash`, `gemini-3.7-flash`, `openai/gpt-live-1`
 and the two `grok-voice-*` realtime entries do not — such an agent is told it
@@ -1291,9 +1298,22 @@ recordings and cost.
 
 ## Email outbound
 
-To email a *list* rather than one person, create an **email batch**: a CSV, an
-agent task that drafts each row, a connected Resend account, and a mapping that
-says which column carries the address, the subject and the body.
+To email a *list* rather than one person, create an **email batch**: a CSV, a
+connected Resend account, and one of two writers for the emails:
+
+- **A template** — `template: {to, subject, body}`, the same email for every
+  row. `to` names the address column; `{{column}}` in the subject or body takes
+  that row's cell (bare column names, no `vars.` prefix), HTML-escaped when
+  `body_format` is `html`. No task, no drafting, no cost: the batch is created
+  `drafted` and every row is ready to review at once. A row whose placeholder
+  cell is blank is `draft_failed` ("first_name is empty") until the cell is
+  filled with `patch_email_batch_recipient`. Before creating one, offer
+  `send_test_email` — it mails the filled template to the signed-in user only.
+- **An agent task** — `task_id` + `field_map`, when each email should be
+  written for its row (research, personalization). The rest of this section
+  is mostly about this kind.
+
+Pick the template when the wording is fixed and only names/fields change.
 
 **Four things to say before creating one.** A user who discovers any of them
 afterwards will not forgive us:
@@ -1303,7 +1323,7 @@ afterwards will not forgive us:
   opted in. This runs on *their* Resend account under *their* agreement.
 - **Creating a batch only drafts.** Nothing is sent until someone creates a
   send — and a send of every row mails drafts nobody reviewed.
-- **Every row runs the task's CURRENT PUBLISHED version**, re-read as drafting
+- **In a task batch, every row runs the task's CURRENT PUBLISHED version**, re-read as drafting
   goes, so publishing mid-batch changes every row drafted after that moment. Which version
   wrote which rows is on the batch as `drafted_versions`, and
   `redraft_email_batch_recipients` with `selection: "stale_version"` is how the

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@talqing/mcp",
-  "version": "0.14.0",
+  "version": "0.16.0",
   "description": "MCP server for building Talqing AI voice, video and text agents from Claude Code, Codex, or any MCP client.",
   "license": "MIT",
   "homepage": "https://talqing.com",

package/tools.json

@@ -1998,7 +1998,28 @@
       "role": "read"
     },
     {
-      "description": "Create an email batch: a list, an agent task that drafts each row, and a connected Resend account to send from.",
+      "description": "Fill a template from `vars` and email it to yourself (the signed-in user) only.",
+      "method": "POST",
+      "name": "send_test_email",
+      "parameters": {
+        "additionalProperties": false,
+        "description": "Arguments not inlined: call describe_function('send_test_email').",
+        "properties": {
+          "body": {
+            "type": "object"
+          }
+        },
+        "required": [
+          "body"
+        ],
+        "type": "object"
+      },
+      "path": "/v1/email/test-sends",
+      "read_only": false,
+      "role": "write"
+    },
+    {
+      "description": "Create an email batch: a list, what writes each email, and a connected Resend account to send from.",
       "method": "POST",
       "name": "create_email_batch",
       "parameters": {