@vocalstack/mcp 1.4.0 → 1.5.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +4 -4
- package/dist/handlers.js +216 -112
- package/dist/tools.json +170 -115
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -62,18 +62,18 @@ Both offer the same tools.
|
|
|
62
62
|
|
|
63
63
|
<!-- tools -->
|
|
64
64
|
- `create_transcription`: Start transcribing a finished recording, from a public link or a file sent inline; returns a transcriptionId at once, before any text.
|
|
65
|
-
- `get_transcription`: Read one transcription's status and text, with timestamps and speakers: as
|
|
65
|
+
- `get_transcription`: Read one transcription's status and text, with timestamps and speakers, or its title and summary: as spoken or translated, at once or after waiting for it to finish.
|
|
66
66
|
- `list_transcriptions`: List this account's transcriptions, newest first, a page at a time: files, links and live ones, including those broadcasts recorded.
|
|
67
67
|
- `search_transcriptions`: Find what was said across this account's transcriptions, or in one: the lines with given words or a name (mode words), or the passages that answer a question (mode meaning).
|
|
68
|
-
- `
|
|
69
|
-
- `rename_speaker`: Name or rename a speaker: one person in one transcription, so its transcript shows the name instead of "Speaker N", or a saved speaker, everywhere they are recognised.
|
|
68
|
+
- `update_transcription`: Change a transcription: its name, its public share link, its summary, or whether spoken translations use the speakers' own cloned voices.
|
|
70
69
|
- `delete_transcription`: Move a transcription to the trash, with its recordings, transcript, translations and summary; its share links stop working.
|
|
71
|
-
- `set_cloned_voices`: Turn cloned voices on or off for spoken translations: for one transcription, or for one saved speaker everywhere they are recognised.
|
|
72
70
|
- `list_speakers`: List the speakers this account has saved: people recognised by name in every new transcription.
|
|
73
71
|
- `save_speaker`: Save one speaker of a transcription to the account's saved speakers, so later transcriptions recognise and name them; their own speech in this transcription is the sample.
|
|
72
|
+
- `update_speaker`: Change a speaker: name one person in one transcription (instead of "Speaker N"), or rename a saved speaker and switch their cloned voice on or off, everywhere they are recognised.
|
|
74
73
|
- `delete_speaker`: Delete a saved speaker for good: their name, fingerprint, stored sample and any cloned voice are erased at once and cannot be restored.
|
|
75
74
|
- `list_broadcasts`: List this account's broadcasts, or get one in detail.
|
|
76
75
|
- `start_broadcast`: Start live transcription and translation of a stream playing now, on a public link listeners open in their own language.
|
|
76
|
+
- `update_broadcast`: Change a broadcast's settings: its name, who can listen (public, or private with a password), the language spoken, and whether translations are read aloud, in stock or the speakers' own cloned voices.
|
|
77
77
|
- `stop_broadcast`: Stop the live transcription a broadcast is running, which also stops its billing.
|
|
78
78
|
- `delete_broadcast`: Move a broadcast to the trash: its public link and stream addresses stop working for listeners.
|
|
79
79
|
<!-- /tools -->
|
package/dist/handlers.js
CHANGED
|
@@ -472,7 +472,7 @@ async function translatedLines(ctx, id, snap, t, lang, offset, limit) {
|
|
|
472
472
|
async function getTranscription(a, ctx) {
|
|
473
473
|
const id = a.id("transcriptionId");
|
|
474
474
|
const waitS = a.int("waitS", 0, 0, WAIT_MAX_S);
|
|
475
|
-
const format = a.oneOf("format", ["text", "segments", "none"], "text");
|
|
475
|
+
const format = a.oneOf("format", ["text", "segments", "summary", "none"], "text");
|
|
476
476
|
const lang = await targetLang(a, ctx, "lang");
|
|
477
477
|
const words = a.bool("includeWords");
|
|
478
478
|
if (words && lang !== undefined)
|
|
@@ -501,6 +501,12 @@ async function getTranscription(a, ctx) {
|
|
|
501
501
|
out.timedOut = timedOut;
|
|
502
502
|
if (format === "none")
|
|
503
503
|
return out;
|
|
504
|
+
if (format === "summary") {
|
|
505
|
+
if (words)
|
|
506
|
+
throw bad("`includeWords` goes with format segments");
|
|
507
|
+
out.summary = await summaryOf(ctx, id, t, lang);
|
|
508
|
+
return out;
|
|
509
|
+
}
|
|
504
510
|
if (lang !== undefined) {
|
|
505
511
|
let notYet;
|
|
506
512
|
if (t.job && t.job.state !== "done")
|
|
@@ -624,11 +630,7 @@ async function searchMeaning(a, ctx) {
|
|
|
624
630
|
async function search(a, ctx) {
|
|
625
631
|
return a.oneOf("mode", ["words", "meaning"], "words") === "meaning" ? searchMeaning(a, ctx) : searchWords(a, ctx);
|
|
626
632
|
}
|
|
627
|
-
async function
|
|
628
|
-
const id = a.id("transcriptionId");
|
|
629
|
-
const lang = await targetLang(a, ctx, "lang");
|
|
630
|
-
const t = await transcription(ctx, id);
|
|
631
|
-
const pending = (message) => ({ state: "pending", message });
|
|
633
|
+
async function summaryOf(ctx, id, t, lang) {
|
|
632
634
|
const s = await ctx.api.get(path("/v1/transcriptions/", id, "/summary"), { lang }, ctx.signal);
|
|
633
635
|
if (s.state === "ready") {
|
|
634
636
|
const v = { state: "ready", title: s.title ?? t.name, summary: s.summary };
|
|
@@ -639,20 +641,28 @@ async function summary(a, ctx) {
|
|
|
639
641
|
return { state: "unavailable", message: "The transcription failed, so there is nothing to summarise." };
|
|
640
642
|
}
|
|
641
643
|
if ((t.job && !TERMINAL.has(t.job.state)) || (!t.job && OPEN.has(t.state))) {
|
|
642
|
-
return pending
|
|
644
|
+
return { state: "pending", message: "The transcription is still running; wait for it (get_transcription with waitS), then ask again." };
|
|
643
645
|
}
|
|
644
646
|
if (s.state === "pending")
|
|
645
|
-
return pending
|
|
646
|
-
|
|
647
|
-
|
|
648
|
-
|
|
649
|
-
|
|
650
|
-
|
|
651
|
-
|
|
652
|
-
|
|
653
|
-
|
|
654
|
-
|
|
647
|
+
return { state: "pending", message: "The summary is being written; ask again in a few seconds." };
|
|
648
|
+
if (s.state === "unavailable")
|
|
649
|
+
return { state: "unavailable", message: "Summaries are not available on this server." };
|
|
650
|
+
const length = t.durationS ?? t.job?.source?.durationS;
|
|
651
|
+
if (typeof length === "number" && length < 60) {
|
|
652
|
+
return { state: "not_requested", message: "Transcriptions shorter than a minute are not summarised." };
|
|
653
|
+
}
|
|
654
|
+
return { state: "none", message: "No summary was asked for; update_transcription with summary true writes one." };
|
|
655
|
+
}
|
|
656
|
+
async function writeSummary(ctx, id, t) {
|
|
657
|
+
const current = await summaryOf(ctx, id, t);
|
|
658
|
+
if (current.state === "ready" || current.state === "unavailable" || current.state === "not_requested")
|
|
659
|
+
return current;
|
|
660
|
+
if ((t.job && !TERMINAL.has(t.job.state)) || (!t.job && OPEN.has(t.state))) {
|
|
661
|
+
throw new ToolError("invalid_state", "The transcription is still running; wait for it (get_transcription with waitS), then ask for the summary.");
|
|
655
662
|
}
|
|
663
|
+
if (current.state === "pending")
|
|
664
|
+
return current;
|
|
665
|
+
const made = (await ctx.api.request("POST", path("/v1/transcriptions/", id, "/summarize"), { signal: ctx.signal })).data;
|
|
656
666
|
switch (made.state) {
|
|
657
667
|
case "ready":
|
|
658
668
|
return { state: "ready", title: made.title ?? t.name, summary: made.summary };
|
|
@@ -661,7 +671,7 @@ async function summary(a, ctx) {
|
|
|
661
671
|
case "not_requested":
|
|
662
672
|
return { state: "not_requested", message: "Transcriptions shorter than a minute are not summarised." };
|
|
663
673
|
default:
|
|
664
|
-
return pending
|
|
674
|
+
return { state: "pending", message: "The summary is being written; ask again in a few seconds." };
|
|
665
675
|
}
|
|
666
676
|
}
|
|
667
677
|
async function all(ctx, route, field, query = {}) {
|
|
@@ -812,7 +822,15 @@ async function deleteTranscription(a, ctx) {
|
|
|
812
822
|
put(out, "restorableUntil", t.purgeAfter);
|
|
813
823
|
return out;
|
|
814
824
|
}
|
|
815
|
-
function
|
|
825
|
+
function optBool(a, k) {
|
|
826
|
+
const v = a.raw[k];
|
|
827
|
+
if (v === undefined || v === null)
|
|
828
|
+
return undefined;
|
|
829
|
+
if (typeof v !== "boolean")
|
|
830
|
+
throw bad(`\`${k}\` must be true or false`);
|
|
831
|
+
return v;
|
|
832
|
+
}
|
|
833
|
+
function speakerTarget(a) {
|
|
816
834
|
const sid = present(a, "speakerId");
|
|
817
835
|
const tid = present(a, "transcriptionId");
|
|
818
836
|
const spk = present(a, "speaker");
|
|
@@ -834,66 +852,6 @@ function noSavedSpeaker(e) {
|
|
|
834
852
|
}
|
|
835
853
|
throw e;
|
|
836
854
|
}
|
|
837
|
-
async function renameSavedSpeaker(a, ctx) {
|
|
838
|
-
const id = a.id("speakerId");
|
|
839
|
-
const given = a.raw.name;
|
|
840
|
-
if (typeof given !== "string")
|
|
841
|
-
throw bad("`name` must be a string");
|
|
842
|
-
const name = given.trim();
|
|
843
|
-
const n = [...name].length;
|
|
844
|
-
if (n === 0 || n > 60)
|
|
845
|
-
throw bad("a saved speaker's `name` must be 1 to 60 characters");
|
|
846
|
-
let s;
|
|
847
|
-
try {
|
|
848
|
-
s = (await ctx.api.request("PATCH", path("/v1/speakers/", id), { json: { name }, signal: ctx.signal })).data;
|
|
849
|
-
}
|
|
850
|
-
catch (e) {
|
|
851
|
-
noSavedSpeaker(e);
|
|
852
|
-
}
|
|
853
|
-
return { speakerId: s.id, name: s.name };
|
|
854
|
-
}
|
|
855
|
-
async function renameSpeaker(a, ctx) {
|
|
856
|
-
if (renameTarget(a))
|
|
857
|
-
return renameSavedSpeaker(a, ctx);
|
|
858
|
-
const id = a.id("transcriptionId");
|
|
859
|
-
const number = a.int("speaker", 1, 1, 1000);
|
|
860
|
-
const given = a.raw.name;
|
|
861
|
-
if (typeof given !== "string")
|
|
862
|
-
throw bad('`name` must be a string ("" clears the name)');
|
|
863
|
-
const name = [...given.trim()].slice(0, 80).join("");
|
|
864
|
-
await transcription(ctx, id);
|
|
865
|
-
try {
|
|
866
|
-
await ctx.api.request("PUT", path("/v1/transcriptions/", id, "/appearances/", String(number)), {
|
|
867
|
-
json: { name: name || null },
|
|
868
|
-
signal: ctx.signal,
|
|
869
|
-
});
|
|
870
|
-
}
|
|
871
|
-
catch (e) {
|
|
872
|
-
if (e instanceof ToolError && e.code === "not_found") {
|
|
873
|
-
throw new ToolError("not_found", `This transcription has no Speaker ${number}; get_transcription shows the speaker numbers.`);
|
|
874
|
-
}
|
|
875
|
-
throw e;
|
|
876
|
-
}
|
|
877
|
-
return { transcriptionId: id, speaker: number, name: name || null };
|
|
878
|
-
}
|
|
879
|
-
const VOICE_CLONING_NEEDS_CONSENT = "turning voiceCloning on needs voiceCloningConsent: true — confirm you have permission from everyone speaking";
|
|
880
|
-
function clonedVoicesTarget(a) {
|
|
881
|
-
const sid = present(a, "speakerId");
|
|
882
|
-
const tid = present(a, "transcriptionId");
|
|
883
|
-
if (sid && !tid)
|
|
884
|
-
return true;
|
|
885
|
-
if (tid && !sid)
|
|
886
|
-
return false;
|
|
887
|
-
if (sid)
|
|
888
|
-
throw bad("give either `transcriptionId` (one transcription) or `speakerId` (a saved speaker), not both");
|
|
889
|
-
throw bad("give `transcriptionId` (one transcription) or `speakerId` (a saved speaker)");
|
|
890
|
-
}
|
|
891
|
-
function onArg(a) {
|
|
892
|
-
const on = a.raw.on;
|
|
893
|
-
if (typeof on !== "boolean")
|
|
894
|
-
throw bad("`on` must be true or false");
|
|
895
|
-
return on;
|
|
896
|
-
}
|
|
897
855
|
async function savedSpeaker(ctx, id) {
|
|
898
856
|
try {
|
|
899
857
|
return await ctx.api.get(path("/v1/speakers/", id), undefined, ctx.signal);
|
|
@@ -910,18 +868,51 @@ async function patchSpeaker(ctx, id, body) {
|
|
|
910
868
|
noSavedSpeaker(e);
|
|
911
869
|
}
|
|
912
870
|
}
|
|
913
|
-
async function
|
|
871
|
+
async function updateSavedSpeaker(a, ctx) {
|
|
914
872
|
const id = a.id("speakerId");
|
|
915
|
-
|
|
873
|
+
let name;
|
|
874
|
+
const given = a.raw.name;
|
|
875
|
+
if (given !== undefined && given !== null) {
|
|
876
|
+
if (typeof given !== "string")
|
|
877
|
+
throw bad("`name` must be a string");
|
|
878
|
+
name = given.trim();
|
|
879
|
+
const n = [...name].length;
|
|
880
|
+
if (n === 0 || n > 60)
|
|
881
|
+
throw bad("a saved speaker's `name` must be 1 to 60 characters");
|
|
882
|
+
}
|
|
883
|
+
const cloned = optBool(a, "clonedVoice");
|
|
916
884
|
const consent = a.bool("consent");
|
|
917
|
-
if (
|
|
885
|
+
if (name === undefined && cloned === undefined)
|
|
886
|
+
throw bad("give `name`, `clonedVoice` or both to change a saved speaker");
|
|
887
|
+
if (cloned === true && !consent) {
|
|
918
888
|
throw bad("Turning a saved speaker's cloned voice on needs consent: true, confirming that person agreed to have their voice cloned.");
|
|
919
889
|
}
|
|
920
|
-
const
|
|
921
|
-
|
|
922
|
-
|
|
923
|
-
|
|
890
|
+
const report = (s) => {
|
|
891
|
+
const out = {
|
|
892
|
+
speakerId: s.id,
|
|
893
|
+
name: s.name,
|
|
894
|
+
consent: s.consent,
|
|
895
|
+
clonedVoice: s.voice?.enabled === true && s.consent === "granted",
|
|
896
|
+
};
|
|
897
|
+
if (s.voice)
|
|
898
|
+
put(out, "voiceState", s.voice.state);
|
|
899
|
+
return out;
|
|
900
|
+
};
|
|
901
|
+
if (cloned === undefined) {
|
|
902
|
+
try {
|
|
903
|
+
return report((await ctx.api.request("PATCH", path("/v1/speakers/", id), { json: { name }, signal: ctx.signal })).data);
|
|
904
|
+
}
|
|
905
|
+
catch (e) {
|
|
906
|
+
noSavedSpeaker(e);
|
|
924
907
|
}
|
|
908
|
+
}
|
|
909
|
+
const cur = await savedSpeaker(ctx, id);
|
|
910
|
+
if (cloned === true && cur.consent === "refused") {
|
|
911
|
+
throw new ToolError("invalid_state", "This speaker refused to be cloned and their audio was deleted; save them again from a transcription (save_speaker with consent) to clone them.");
|
|
912
|
+
}
|
|
913
|
+
if (name !== undefined)
|
|
914
|
+
await patchSpeaker(ctx, id, { name });
|
|
915
|
+
if (cloned === true) {
|
|
925
916
|
let s;
|
|
926
917
|
try {
|
|
927
918
|
s = (await ctx.api.request("POST", path("/v1/speakers/", id, "/consent"), { json: { consent: "granted" }, signal: ctx.signal })).data;
|
|
@@ -935,41 +926,154 @@ async function setSpeakerClonedVoice(a, ctx) {
|
|
|
935
926
|
if (s?.voice && s.voice.enabled === false)
|
|
936
927
|
await patchSpeaker(ctx, id, { enabled: true });
|
|
937
928
|
}
|
|
938
|
-
else if (cur.voice?.enabled === true) {
|
|
929
|
+
else if (cloned === false && cur.voice?.enabled === true) {
|
|
939
930
|
await patchSpeaker(ctx, id, { enabled: false });
|
|
940
931
|
}
|
|
941
|
-
|
|
942
|
-
const out = {
|
|
943
|
-
speakerId: s.id,
|
|
944
|
-
name: s.name,
|
|
945
|
-
consent: s.consent,
|
|
946
|
-
clonedVoice: s.voice?.enabled === true && s.consent === "granted",
|
|
947
|
-
};
|
|
948
|
-
if (s.voice)
|
|
949
|
-
put(out, "voiceState", s.voice.state);
|
|
950
|
-
return out;
|
|
932
|
+
return report(await savedSpeaker(ctx, id));
|
|
951
933
|
}
|
|
952
|
-
async function
|
|
953
|
-
if (
|
|
954
|
-
return
|
|
934
|
+
async function updateSpeaker(a, ctx) {
|
|
935
|
+
if (speakerTarget(a))
|
|
936
|
+
return updateSavedSpeaker(a, ctx);
|
|
937
|
+
if (present(a, "clonedVoice")) {
|
|
938
|
+
throw bad("`clonedVoice` is for a saved speaker (speakerId); for everyone in one transcription use update_transcription with clonedVoices");
|
|
939
|
+
}
|
|
955
940
|
const id = a.id("transcriptionId");
|
|
956
|
-
const
|
|
957
|
-
const
|
|
941
|
+
const number = a.int("speaker", 1, 1, 1000);
|
|
942
|
+
const given = a.raw.name;
|
|
943
|
+
if (given === undefined || given === null)
|
|
944
|
+
throw bad('`name` is required with `transcriptionId` and `speaker` ("" clears the name)');
|
|
945
|
+
if (typeof given !== "string")
|
|
946
|
+
throw bad('`name` must be a string ("" clears the name)');
|
|
947
|
+
const name = [...given.trim()].slice(0, 80).join("");
|
|
958
948
|
await transcription(ctx, id);
|
|
959
|
-
let t;
|
|
960
949
|
try {
|
|
961
|
-
|
|
962
|
-
json: {
|
|
950
|
+
await ctx.api.request("PUT", path("/v1/transcriptions/", id, "/appearances/", String(number)), {
|
|
951
|
+
json: { name: name || null },
|
|
963
952
|
signal: ctx.signal,
|
|
964
|
-
})
|
|
953
|
+
});
|
|
965
954
|
}
|
|
966
955
|
catch (e) {
|
|
967
|
-
if (e instanceof ToolError && e.
|
|
968
|
-
throw
|
|
956
|
+
if (e instanceof ToolError && e.code === "not_found") {
|
|
957
|
+
throw new ToolError("not_found", `This transcription has no Speaker ${number}; get_transcription shows the speaker numbers.`);
|
|
969
958
|
}
|
|
970
959
|
throw e;
|
|
971
960
|
}
|
|
972
|
-
return { transcriptionId:
|
|
961
|
+
return { transcriptionId: id, speaker: number, name: name || null };
|
|
962
|
+
}
|
|
963
|
+
const VOICE_CLONING_NEEDS_CONSENT = "turning voiceCloning on needs voiceCloningConsent: true — confirm you have permission from everyone speaking";
|
|
964
|
+
const NEEDS_CONSENT = "Turning cloned voices on needs consent: true, confirming everyone speaking agreed to have their voice cloned.";
|
|
965
|
+
function consentError(e) {
|
|
966
|
+
if (e instanceof ToolError && e.message === VOICE_CLONING_NEEDS_CONSENT)
|
|
967
|
+
throw bad(NEEDS_CONSENT);
|
|
968
|
+
throw e;
|
|
969
|
+
}
|
|
970
|
+
async function updateTranscription(a, ctx) {
|
|
971
|
+
const id = a.id("transcriptionId");
|
|
972
|
+
let name = a.str("name");
|
|
973
|
+
if (name !== undefined) {
|
|
974
|
+
name = name.trim();
|
|
975
|
+
const n = [...name].length;
|
|
976
|
+
if (n === 0 || n > 200)
|
|
977
|
+
throw bad("`name` must be 1 to 200 characters");
|
|
978
|
+
}
|
|
979
|
+
const share = optBool(a, "shareLink");
|
|
980
|
+
const summary = optBool(a, "summary");
|
|
981
|
+
const cloned = optBool(a, "clonedVoices");
|
|
982
|
+
const consent = a.bool("consent");
|
|
983
|
+
if (summary === false)
|
|
984
|
+
throw bad("`summary` takes only true (write one now); a summary is never deleted");
|
|
985
|
+
if (name === undefined && share === undefined && summary === undefined && cloned === undefined) {
|
|
986
|
+
throw bad("give at least one of `name`, `shareLink`, `summary` or `clonedVoices` to change");
|
|
987
|
+
}
|
|
988
|
+
if (cloned === true && !consent)
|
|
989
|
+
throw bad(NEEDS_CONSENT);
|
|
990
|
+
let t = await transcription(ctx, id);
|
|
991
|
+
if (t.lifecycle && t.lifecycle !== "active") {
|
|
992
|
+
throw new ToolError("invalid_state", "This transcription is in the trash; restore it in the VocalStack Dashboard first.");
|
|
993
|
+
}
|
|
994
|
+
const written = summary === true ? await writeSummary(ctx, id, t) : undefined;
|
|
995
|
+
if (name !== undefined || cloned !== undefined) {
|
|
996
|
+
const body = {};
|
|
997
|
+
if (name !== undefined)
|
|
998
|
+
body.name = name;
|
|
999
|
+
if (cloned !== undefined)
|
|
1000
|
+
body.settings = { voiceCloning: cloned, voiceCloningConsent: consent };
|
|
1001
|
+
try {
|
|
1002
|
+
t = (await ctx.api.request("PATCH", path("/v1/transcriptions/", id), { json: body, signal: ctx.signal })).data;
|
|
1003
|
+
}
|
|
1004
|
+
catch (e) {
|
|
1005
|
+
consentError(e);
|
|
1006
|
+
}
|
|
1007
|
+
}
|
|
1008
|
+
let links = (await ctx.api.get(path("/v1/transcriptions/", id, "/share-links"), undefined, ctx.signal))?.links ?? [];
|
|
1009
|
+
if (share === true && links.length === 0) {
|
|
1010
|
+
const l = (await ctx.api.request("POST", path("/v1/transcriptions/", id, "/share-links"), { json: { label: "Shared link" }, signal: ctx.signal })).data;
|
|
1011
|
+
links = [l];
|
|
1012
|
+
}
|
|
1013
|
+
else if (share === false) {
|
|
1014
|
+
for (const l of links)
|
|
1015
|
+
await ctx.api.request("DELETE", path("/v1/share-links/", l.id), { signal: ctx.signal });
|
|
1016
|
+
links = [];
|
|
1017
|
+
}
|
|
1018
|
+
const out = { transcriptionId: t.id, name: t.name, shareUrl: links[0]?.url ?? null, clonedVoices: t.settings?.voiceCloning === true };
|
|
1019
|
+
if (written)
|
|
1020
|
+
out.summary = written;
|
|
1021
|
+
return out;
|
|
1022
|
+
}
|
|
1023
|
+
async function updateBroadcast(a, ctx) {
|
|
1024
|
+
const id = a.id("broadcastId");
|
|
1025
|
+
const name = a.str("name");
|
|
1026
|
+
const access = a.raw.access === undefined || a.raw.access === null ? undefined : a.oneOf("access", ["public", "private"], "public");
|
|
1027
|
+
const password = a.str("password");
|
|
1028
|
+
const spokenRaw = a.str("spokenLanguage");
|
|
1029
|
+
const speak = optBool(a, "spokenTranslations");
|
|
1030
|
+
const cloned = optBool(a, "clonedVoices");
|
|
1031
|
+
const consent = a.bool("consent");
|
|
1032
|
+
if ([name, access, password, spokenRaw, speak, cloned].every((v) => v === undefined)) {
|
|
1033
|
+
throw bad("give at least one of `name`, `access`, `password`, `spokenLanguage`, `spokenTranslations` or `clonedVoices` to change");
|
|
1034
|
+
}
|
|
1035
|
+
if (cloned === true && !consent)
|
|
1036
|
+
throw bad(NEEDS_CONSENT);
|
|
1037
|
+
await broadcast(ctx, id);
|
|
1038
|
+
const body = {};
|
|
1039
|
+
if (name !== undefined)
|
|
1040
|
+
body.name = name;
|
|
1041
|
+
if (access !== undefined)
|
|
1042
|
+
body.access = access;
|
|
1043
|
+
if (password !== undefined && access !== "public")
|
|
1044
|
+
body.password = password;
|
|
1045
|
+
const settings = {};
|
|
1046
|
+
if (spokenRaw !== undefined) {
|
|
1047
|
+
const l = spokenRaw.trim().toLowerCase();
|
|
1048
|
+
settings.spokenLanguage = l === "" || l === "auto" ? null : l;
|
|
1049
|
+
}
|
|
1050
|
+
if (speak !== undefined)
|
|
1051
|
+
settings.translationTts = speak;
|
|
1052
|
+
if (cloned !== undefined) {
|
|
1053
|
+
settings.voiceCloning = cloned;
|
|
1054
|
+
settings.voiceCloningConsent = consent;
|
|
1055
|
+
}
|
|
1056
|
+
if (Object.keys(settings).length)
|
|
1057
|
+
body.settings = settings;
|
|
1058
|
+
let b;
|
|
1059
|
+
try {
|
|
1060
|
+
b = (await ctx.api.request("PATCH", path("/v1/broadcasts/", id), { json: body, signal: ctx.signal })).data;
|
|
1061
|
+
}
|
|
1062
|
+
catch (e) {
|
|
1063
|
+
if (e instanceof ToolError && e.code === "invalid_state") {
|
|
1064
|
+
throw new ToolError("invalid_state", "This broadcast is in the trash; restore it in the VocalStack Dashboard first.");
|
|
1065
|
+
}
|
|
1066
|
+
if (e instanceof ToolError && e.message === "private broadcasts need a password") {
|
|
1067
|
+
throw bad("A private broadcast needs a `password` (at least 4 characters) the first time.");
|
|
1068
|
+
}
|
|
1069
|
+
consentError(e);
|
|
1070
|
+
}
|
|
1071
|
+
return {
|
|
1072
|
+
...brief(b),
|
|
1073
|
+
spokenLanguage: b.settings?.spokenLanguage ?? "auto",
|
|
1074
|
+
spokenTranslations: b.settings?.translationTts === true,
|
|
1075
|
+
clonedVoices: b.settings?.voiceCloning === true,
|
|
1076
|
+
};
|
|
973
1077
|
}
|
|
974
1078
|
const ON_AIR = "This broadcast is already on air; stop it first (stop_broadcast) or use another broadcast.";
|
|
975
1079
|
async function startBroadcast(a, ctx) {
|
|
@@ -1084,15 +1188,15 @@ export const handlers = {
|
|
|
1084
1188
|
get_transcription: getTranscription,
|
|
1085
1189
|
list_transcriptions: listTranscriptions,
|
|
1086
1190
|
search_transcriptions: search,
|
|
1087
|
-
|
|
1088
|
-
rename_speaker: renameSpeaker,
|
|
1191
|
+
update_transcription: updateTranscription,
|
|
1089
1192
|
delete_transcription: deleteTranscription,
|
|
1090
|
-
set_cloned_voices: setClonedVoices,
|
|
1091
1193
|
list_speakers: speakers,
|
|
1092
1194
|
save_speaker: saveSpeaker,
|
|
1195
|
+
update_speaker: updateSpeaker,
|
|
1093
1196
|
delete_speaker: deleteSpeaker,
|
|
1094
1197
|
list_broadcasts: listBroadcasts,
|
|
1095
1198
|
start_broadcast: startBroadcast,
|
|
1199
|
+
update_broadcast: updateBroadcast,
|
|
1096
1200
|
stop_broadcast: stopBroadcast,
|
|
1097
1201
|
delete_broadcast: deleteBroadcast,
|
|
1098
1202
|
};
|
package/dist/tools.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
|
1
1
|
{
|
|
2
|
-
"instructions": "VocalStack transcribes and translates recorded or live audio and video in about 100 languages, with speaker labels.\nRecorded audio: create_transcription (a link or a file) returns a transcriptionId at once; get_transcription with waitS waits for it and reads the text, in any language with lang.\nFind: list_transcriptions (ids, newest first); search_transcriptions (mode words for words or names, meaning for a question).\nOne transcription:
|
|
2
|
+
"instructions": "VocalStack transcribes and translates recorded or live audio and video in about 100 languages, with speaker labels.\nRecorded audio: create_transcription (a link or a file) returns a transcriptionId at once; get_transcription with waitS waits for it and reads the text, in any language with lang.\nFind: list_transcriptions (ids, newest first); search_transcriptions (mode words for words or names, meaning for a question).\nOne transcription: get_transcription with format summary for its title and summary; update_transcription (name, share link, summary, cloned voices); delete_transcription (to the trash, restorable in the Dashboard).\nLive streams: start_broadcast, then get_transcription while it runs, stop_broadcast to end it; list_broadcasts shows them; update_broadcast changes access, language and spoken translations; delete_broadcast removes one.\nPeople: list_speakers, save_speaker (voice cloned only with the person's consent), update_speaker (names, and a saved speaker's cloned voice), delete_speaker.\nOnly transcription is billed; everything else, and listening, is free. Retry a start with the same idempotencyKey.\nErrors are results with isError true: error, message, retryable, and retryAfterS or topUpUrl when they apply.",
|
|
3
3
|
"tools": [
|
|
4
4
|
{
|
|
5
5
|
"name": "create_transcription",
|
|
@@ -84,7 +84,7 @@
|
|
|
84
84
|
},
|
|
85
85
|
"summary": {
|
|
86
86
|
"default": false,
|
|
87
|
-
"description": "Also write a title and summary when it is done, at no extra cost (
|
|
87
|
+
"description": "Also write a title and summary when it is done, at no extra cost (get_transcription with format summary reads it). Never for recordings under a minute.",
|
|
88
88
|
"type": "boolean"
|
|
89
89
|
},
|
|
90
90
|
"url": {
|
|
@@ -111,16 +111,17 @@
|
|
|
111
111
|
{
|
|
112
112
|
"name": "get_transcription",
|
|
113
113
|
"title": "Read a transcription",
|
|
114
|
-
"description": "Read one transcription's status and text, with timestamps and speakers: as
|
|
114
|
+
"description": "Read one transcription's status and text, with timestamps and speakers, or its title and summary: as spoken or translated, at once or after waiting for it to finish.\nUse: after create_transcription (waitS saves polling), or any time during a live transcription for the text so far. An overview: format summary. Finding a passage in many transcriptions: search_transcriptions.\nBehaviour: free. A wait ends as soon as the work does, and sends progress notifications when the client gives a progress token. lang needs a finished transcription; the first call in a new language can take up to a minute on a long one, later pages are quick. Page with offset = nextOffset.\nReturns: status (state, progress, error when failed, shareUrl when it has a public link: made with shareLink in create_transcription or update_transcription), timedOut when waitS was set, then segmentCount, partial while text is still coming, text or segments, nextOffset when more remain, and lang and translated with lang. format summary returns summary instead: state ready with title and summary (and lang when translated), pending while it is written or the transcription runs, none when none was asked for (update_transcription with summary true writes one), not_requested under a minute, unavailable when it failed. A wait that runs out with lang set returns the status only.\nErrors: not_found; invalid_state for lang before it is finished; invalid_options for an unsupported lang or includeWords with lang; rate_limited when waitS is set and 4 waits are already open for this key (8 for the account).",
|
|
115
115
|
"inputSchema": {
|
|
116
116
|
"additionalProperties": false,
|
|
117
117
|
"properties": {
|
|
118
118
|
"format": {
|
|
119
119
|
"default": "text",
|
|
120
|
-
"description": "text: one string, a line per segment as \"[mm:ss] Speaker 1: words\", for reading and quoting; segments: timed lines (start, end, text, speakerNumber, speakerName), for subtitles and exact times; none: status only (with waitS: wait without reading).",
|
|
120
|
+
"description": "text: one string, a line per segment as \"[mm:ss] Speaker 1: words\", for reading and quoting; segments: timed lines (start, end, text, speakerNumber, speakerName), for subtitles and exact times; summary: the title and summary only, for an overview; none: status only (with waitS: wait without reading).",
|
|
121
121
|
"enum": [
|
|
122
122
|
"text",
|
|
123
123
|
"segments",
|
|
124
|
+
"summary",
|
|
124
125
|
"none"
|
|
125
126
|
]
|
|
126
127
|
},
|
|
@@ -130,7 +131,7 @@
|
|
|
130
131
|
"type": "boolean"
|
|
131
132
|
},
|
|
132
133
|
"lang": {
|
|
133
|
-
"description": "Code of a language to read the whole text in, one of about 100 (e.g. es, de, ja): the transcription must be finished (state done, or a stopped live one). The language already spoken returns the original with translated false. Omit for the original.",
|
|
134
|
+
"description": "Code of a language to read the whole text (or, with format summary, the summary) in, one of about 100 (e.g. es, de, ja): the transcription must be finished (state done, or a stopped live one). The language already spoken returns the original with translated false. Omit for the original.",
|
|
134
135
|
"examples": [
|
|
135
136
|
"es",
|
|
136
137
|
"ja"
|
|
@@ -232,7 +233,7 @@
|
|
|
232
233
|
{
|
|
233
234
|
"name": "search_transcriptions",
|
|
234
235
|
"title": "Search transcriptions",
|
|
235
|
-
"description": "Find what was said across this account's transcriptions, or in one: the lines with given words or a name (mode words), or the passages that answer a question (mode meaning).\nUse: words when you know what was said; meaning for a question or topic, in any language. Text around a hit: get_transcription with offset. One transcription's overview:
|
|
236
|
+
"description": "Find what was said across this account's transcriptions, or in one: the lines with given words or a name (mode words), or the passages that answer a question (mode meaning).\nUse: words when you know what was said; meaning for a question or topic, in any language. Text around a hit: get_transcription with offset. One transcription's overview: get_transcription with format summary.\nBehaviour: free; no paging. words ranks whole words first, then close spellings, then newest, and searches the query's English translation too. meaning returns up to 10 passages, closest first. Searches by meaning an hour per account: 60 on Pay as you go, no limit on Premium.\nReturns: words: query, queryEn, hits (transcriptionId, name, mediaStart in seconds, speaker, text, score, live, recordedAt, rendered with lang). meaning: query, state (ready, or unavailable or off when it cannot run), passages (transcriptionId, name, mediaStart, mediaEnd, text, similarity 0 to 1, live), searchesLeftThisHour when capped. Cite the transcription and second. No match: an empty list.\nErrors: not_found for an unknown transcriptionId; invalid_options for lang or limit with mode meaning; rate_limited (with retryAfterS) past the hourly searches by meaning.",
|
|
236
237
|
"inputSchema": {
|
|
237
238
|
"additionalProperties": false,
|
|
238
239
|
"properties": {
|
|
@@ -295,75 +296,40 @@
|
|
|
295
296
|
}
|
|
296
297
|
},
|
|
297
298
|
{
|
|
298
|
-
"name": "
|
|
299
|
-
"title": "
|
|
300
|
-
"description": "
|
|
299
|
+
"name": "update_transcription",
|
|
300
|
+
"title": "Update a transcription",
|
|
301
|
+
"description": "Change a transcription: its name, its public share link, its summary, or whether spoken translations use the speakers' own cloned voices.\nUse: renaming, sharing or unsharing, asking for a summary later, or switching cloned voices. Only the fields given change; give at least one. Speaker names: update_speaker. Removing it: delete_transcription.\nArguments: transcriptionId (required). name 1 to 200 characters. shareLink true makes a public read-only link (or keeps the one it has), false turns every public link off. summary true writes a title and summary if it has none (a finished transcription of at least a minute). clonedVoices true needs consent true, recording that everyone speaking agreed to be cloned; false needs no consent.\nBehaviour: free; calling again with the same fields changes nothing. A link turned off cannot be turned back on: shareLink true afterwards makes a new address. With cloned voices on, each speaker's translated words are read in that speaker's own voice, never anyone else's; speakers who refused, or without a usable clone, keep a stock voice; lines read after the change use it.\nReturns: transcriptionId, name, shareUrl (null with no public link), summary (state ready, pending, not_requested or unavailable) when asked for, clonedVoices.\nErrors: invalid_options for no field to change, a bad name, summary false, or clonedVoices true without consent; invalid_state for a deleted transcription, or summary before it is finished; not_found; forbidden for a read-only key.",
|
|
301
302
|
"inputSchema": {
|
|
302
303
|
"additionalProperties": false,
|
|
303
304
|
"properties": {
|
|
304
|
-
"
|
|
305
|
-
"description": "
|
|
306
|
-
"
|
|
307
|
-
|
|
308
|
-
|
|
309
|
-
"
|
|
305
|
+
"clonedVoices": {
|
|
306
|
+
"description": "true: translations of each speaker's own words are read aloud in that speaker's own cloned voice (needs consent true); false: stock voices.",
|
|
307
|
+
"type": "boolean"
|
|
308
|
+
},
|
|
309
|
+
"consent": {
|
|
310
|
+
"default": false,
|
|
311
|
+
"description": "Must be true with clonedVoices true: confirms everyone speaking in it agreed to have their voice cloned. Recorded with your account and the time. Not needed to turn cloned voices off.",
|
|
312
|
+
"type": "boolean"
|
|
310
313
|
},
|
|
311
|
-
"transcriptionId": {
|
|
312
|
-
"description": "A transcriptionId from create_transcription, list_transcriptions, search_transcriptions or start_broadcast.",
|
|
313
|
-
"examples": [
|
|
314
|
-
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
315
|
-
],
|
|
316
|
-
"format": "uuid",
|
|
317
|
-
"type": "string"
|
|
318
|
-
}
|
|
319
|
-
},
|
|
320
|
-
"required": [
|
|
321
|
-
"transcriptionId"
|
|
322
|
-
],
|
|
323
|
-
"type": "object"
|
|
324
|
-
},
|
|
325
|
-
"annotations": {
|
|
326
|
-
"title": "Get a summary",
|
|
327
|
-
"readOnlyHint": false,
|
|
328
|
-
"destructiveHint": false,
|
|
329
|
-
"idempotentHint": true,
|
|
330
|
-
"openWorldHint": false
|
|
331
|
-
}
|
|
332
|
-
},
|
|
333
|
-
{
|
|
334
|
-
"name": "rename_speaker",
|
|
335
|
-
"title": "Rename a speaker",
|
|
336
|
-
"description": "Name or rename a speaker: one person in one transcription, so its transcript shows the name instead of \"Speaker N\", or a saved speaker, everywhere they are recognised.\nUse: labelling who spoke here, or correcting a saved speaker's name. To recognise the person in later transcriptions too: save_speaker (it also names them here). Saved speakers: list_speakers.\nArguments: give exactly one of transcriptionId with speaker (the number in \"Speaker 2\" from get_transcription, not a name or an id) or speakerId (from list_speakers). name is trimmed; in a transcription it is cut to 80 characters and an empty name undoes it; a saved speaker's name is 1 to 60 characters. Calling again replaces the name.\nBehaviour: free and reversible; in a transcription it applies to that person in every recording of it, and listeners following it live see it at once. The person's speaker number never changes. Renaming a saved speaker changes nothing else about them (fingerprint, consent, cloned voice).\nReturns: transcriptionId, speaker, name (null when cleared); or, with speakerId, speakerId and name.\nErrors: invalid_options for both or neither kind of target, or a saved speaker's name out of 1 to 60; not_found for an unknown transcription, speaker number or speakerId; forbidden for a read-only key.",
|
|
337
|
-
"inputSchema": {
|
|
338
|
-
"additionalProperties": false,
|
|
339
|
-
"properties": {
|
|
340
314
|
"name": {
|
|
341
|
-
"description": "
|
|
315
|
+
"description": "New display name, shown in lists and on the share page; trimmed, 1 to 200 characters.",
|
|
342
316
|
"examples": [
|
|
343
|
-
"
|
|
344
|
-
""
|
|
317
|
+
"Board meeting, 19 November"
|
|
345
318
|
],
|
|
346
|
-
"maxLength":
|
|
319
|
+
"maxLength": 200,
|
|
320
|
+
"minLength": 1,
|
|
347
321
|
"type": "string"
|
|
348
322
|
},
|
|
349
|
-
"
|
|
350
|
-
"description": "
|
|
351
|
-
"
|
|
352
|
-
2
|
|
353
|
-
],
|
|
354
|
-
"minimum": 1,
|
|
355
|
-
"type": "integer"
|
|
323
|
+
"shareLink": {
|
|
324
|
+
"description": "true: make a public link anyone can open to read the transcript without signing in (an existing one is kept; shareUrl in the result). false: every public link of it stops working at once.",
|
|
325
|
+
"type": "boolean"
|
|
356
326
|
},
|
|
357
|
-
"
|
|
358
|
-
"description": "
|
|
359
|
-
"
|
|
360
|
-
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
361
|
-
],
|
|
362
|
-
"format": "uuid",
|
|
363
|
-
"type": "string"
|
|
327
|
+
"summary": {
|
|
328
|
+
"description": "true: write a title and summary now if it has none (free; the transcription must be finished and at least a minute long). Read it with get_transcription format summary. false is not accepted: a summary is never deleted.",
|
|
329
|
+
"type": "boolean"
|
|
364
330
|
},
|
|
365
331
|
"transcriptionId": {
|
|
366
|
-
"description": "
|
|
332
|
+
"description": "A transcriptionId from create_transcription, list_transcriptions, search_transcriptions or start_broadcast.",
|
|
367
333
|
"examples": [
|
|
368
334
|
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
369
335
|
],
|
|
@@ -372,12 +338,12 @@
|
|
|
372
338
|
}
|
|
373
339
|
},
|
|
374
340
|
"required": [
|
|
375
|
-
"
|
|
341
|
+
"transcriptionId"
|
|
376
342
|
],
|
|
377
343
|
"type": "object"
|
|
378
344
|
},
|
|
379
345
|
"annotations": {
|
|
380
|
-
"title": "
|
|
346
|
+
"title": "Update a transcription",
|
|
381
347
|
"readOnlyHint": false,
|
|
382
348
|
"destructiveHint": false,
|
|
383
349
|
"idempotentHint": true,
|
|
@@ -413,56 +379,10 @@
|
|
|
413
379
|
"openWorldHint": false
|
|
414
380
|
}
|
|
415
381
|
},
|
|
416
|
-
{
|
|
417
|
-
"name": "set_cloned_voices",
|
|
418
|
-
"title": "Turn cloned voices on or off",
|
|
419
|
-
"description": "Turn cloned voices on or off for spoken translations: for one transcription, or for one saved speaker everywhere they are recognised.\nOn: the speaker's translated words are read in that speaker's own cloned voice, never anyone else's; a speaker without a usable clone gets a stock voice. Off: stock voices. Voices are chosen by VocalStack; this is the only voice choice.\nArguments: give exactly one of transcriptionId (every speaker in it) or speakerId (a saved speaker, from list_speakers). on is required; consent must be true to turn on.\nConsent: turning on records, with your account and the time, that everyone speaking (or that saved speaker) agreed. For a transcription it releases its clones waiting for consent; speakers who refused are never cloned. For a saved speaker saved without consent it starts their clone from their stored sample. Off keeps the saved speaker's fingerprint, consent and clone, only unused; to erase the clone and their samples: delete_speaker.\nBehaviour: free and reversible; lines read aloud after the change use the new setting.\nReturns: transcriptionId, name, clonedVoices; or, with speakerId, speakerId, name, consent, clonedVoice, and voiceState (training, then ready) when cloned.\nErrors: invalid_options for on without consent or both or neither target; invalid_state for a deleted transcription, a speaker who refused, or one with no stored sample to clone from; not_found; forbidden for a read-only key.",
|
|
420
|
-
"inputSchema": {
|
|
421
|
-
"additionalProperties": false,
|
|
422
|
-
"properties": {
|
|
423
|
-
"consent": {
|
|
424
|
-
"default": false,
|
|
425
|
-
"description": "Must be true to turn cloned voices on: confirms everyone speaking (with speakerId: that person) agreed to have their voice cloned. Recorded with your account and the time. Not needed to turn them off.",
|
|
426
|
-
"type": "boolean"
|
|
427
|
-
},
|
|
428
|
-
"on": {
|
|
429
|
-
"description": "true: translations of the speaker's own words are read in their own cloned voice; false (the default for a transcription): stock voices.",
|
|
430
|
-
"type": "boolean"
|
|
431
|
-
},
|
|
432
|
-
"speakerId": {
|
|
433
|
-
"description": "Instead of transcriptionId, for one saved speaker everywhere they are recognised: id in list_speakers, or speakerId from save_speaker.",
|
|
434
|
-
"examples": [
|
|
435
|
-
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
436
|
-
],
|
|
437
|
-
"format": "uuid",
|
|
438
|
-
"type": "string"
|
|
439
|
-
},
|
|
440
|
-
"transcriptionId": {
|
|
441
|
-
"description": "For one transcription's spoken translations: a transcriptionId from create_transcription, list_transcriptions, search_transcriptions or start_broadcast. Leave out with speakerId.",
|
|
442
|
-
"examples": [
|
|
443
|
-
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
444
|
-
],
|
|
445
|
-
"format": "uuid",
|
|
446
|
-
"type": "string"
|
|
447
|
-
}
|
|
448
|
-
},
|
|
449
|
-
"required": [
|
|
450
|
-
"on"
|
|
451
|
-
],
|
|
452
|
-
"type": "object"
|
|
453
|
-
},
|
|
454
|
-
"annotations": {
|
|
455
|
-
"title": "Turn cloned voices on or off",
|
|
456
|
-
"readOnlyHint": false,
|
|
457
|
-
"destructiveHint": false,
|
|
458
|
-
"idempotentHint": true,
|
|
459
|
-
"openWorldHint": false
|
|
460
|
-
}
|
|
461
|
-
},
|
|
462
382
|
{
|
|
463
383
|
"name": "list_speakers",
|
|
464
384
|
"title": "List saved speakers",
|
|
465
|
-
"description": "List the speakers this account has saved: people recognised by name in every new transcription.\nUse: to see who is known or find a speakerId for delete_speaker. Save one: save_speaker. Names inside one transcription come from get_transcription and are set with
|
|
385
|
+
"description": "List the speakers this account has saved: people recognised by name in every new transcription.\nUse: to see who is known or find a speakerId for delete_speaker. Save one: save_speaker. Names inside one transcription come from get_transcription and are set with update_speaker, which saves no one.\nBehaviour: free; no paging.\nReturns: speakers, newest first, each with id, name, gender when known, cloned (their voice was cloned, with consent, to read translations of their own words) and, when cloned, voiceReady. None saved: an empty list.",
|
|
466
386
|
"inputSchema": {
|
|
467
387
|
"additionalProperties": false,
|
|
468
388
|
"properties": {},
|
|
@@ -479,7 +399,7 @@
|
|
|
479
399
|
{
|
|
480
400
|
"name": "save_speaker",
|
|
481
401
|
"title": "Save a speaker",
|
|
482
|
-
"description": "Save one speaker of a transcription to the account's saved speakers, so later transcriptions recognise and name them; their own speech in this transcription is the sample.\nUse: remembering a person across transcriptions. This transcription only:
|
|
402
|
+
"description": "Save one speaker of a transcription to the account's saved speakers, so later transcriptions recognise and name them; their own speech in this transcription is the sample.\nUse: remembering a person across transcriptions. This transcription only: update_speaker. Who is saved: list_speakers. Remove: delete_speaker.\nArguments: transcriptionId and speaker point at one person: speaker is the number in \"Speaker 2\" from get_transcription, who must have spoken (a live transcription can still be running). name is 1 to 60 characters, trimmed; renaming later: update_speaker with speakerId. consent is true only when that person agreed to be cloned; leaving it out or false keeps only their fingerprint.\nBehaviour: free; also names them in this transcription. Each call saves a new speaker, so check list_speakers first. Without consent only a voice fingerprint is kept; with it the voice is cloned too, recorded with your account and the time.\nReturns: speakerId, name, consent (granted or pending), cloned, voiceState (training, then ready) when cloned.\nErrors: not_found for an unknown transcription or speaker number; invalid_options for too little speech of them (the message says how much there is); forbidden for a read-only key.",
|
|
483
403
|
"inputSchema": {
|
|
484
404
|
"additionalProperties": false,
|
|
485
405
|
"properties": {
|
|
@@ -529,10 +449,70 @@
|
|
|
529
449
|
"openWorldHint": false
|
|
530
450
|
}
|
|
531
451
|
},
|
|
452
|
+
{
|
|
453
|
+
"name": "update_speaker",
|
|
454
|
+
"title": "Update a speaker",
|
|
455
|
+
"description": "Change a speaker: name one person in one transcription (instead of \"Speaker N\"), or rename a saved speaker and switch their cloned voice on or off, everywhere they are recognised.\nUse: labelling who spoke here, correcting a saved speaker's name, or turning their cloned voice on with their consent. To recognise a person in later transcriptions: save_speaker (it also names them here). Saved speakers: list_speakers.\nArguments: give exactly one of transcriptionId with speaker (the number in \"Speaker 2\" from get_transcription, not a name or an id), with name; or speakerId (from list_speakers) with name, clonedVoice or both. A name in a transcription is cut to 80 characters and an empty one undoes it; a saved speaker's is 1 to 60. clonedVoice true needs consent true.\nBehaviour: free and reversible; calling again replaces the name. In a transcription it applies to that person in every recording of it, and listeners following live see it at once; their speaker number never changes. clonedVoice true clones a saved speaker from their stored sample if they have no clone yet (voiceState training, then ready); false keeps the clone, unused (delete_speaker erases it). Nothing else about the person changes.\nReturns: transcriptionId, speaker, name (null when cleared); or, with speakerId, speakerId, name, consent, clonedVoice, and voiceState when cloned.\nErrors: invalid_options for both or neither kind of target, nothing to change, a bad name, clonedVoice with transcriptionId, or clonedVoice true without consent; invalid_state for a saved speaker who refused cloning or has no sample; not_found for an unknown transcription, speaker number or speakerId; forbidden for a read-only key.",
|
|
456
|
+
"inputSchema": {
|
|
457
|
+
"additionalProperties": false,
|
|
458
|
+
"properties": {
|
|
459
|
+
"clonedVoice": {
|
|
460
|
+
"description": "With speakerId only: true reads translations of this person's own words in their own cloned voice (needs consent true; cloned from their stored sample if not yet), false back to a stock voice (the clone is kept, unused).",
|
|
461
|
+
"type": "boolean"
|
|
462
|
+
},
|
|
463
|
+
"consent": {
|
|
464
|
+
"default": false,
|
|
465
|
+
"description": "Must be true with clonedVoices true: confirms this person agreed to have their voice cloned. Recorded with your account and the time. Not needed to turn cloned voices off.",
|
|
466
|
+
"type": "boolean"
|
|
467
|
+
},
|
|
468
|
+
"name": {
|
|
469
|
+
"description": "The name to show instead of \"Speaker N\", trimmed: up to 80 characters in one transcription (an empty string clears it back to \"Speaker N\"), 1 to 60 for a saved speaker. Required with transcriptionId.",
|
|
470
|
+
"examples": [
|
|
471
|
+
"Maria Lopez",
|
|
472
|
+
""
|
|
473
|
+
],
|
|
474
|
+
"maxLength": 80,
|
|
475
|
+
"type": "string"
|
|
476
|
+
},
|
|
477
|
+
"speaker": {
|
|
478
|
+
"description": "With transcriptionId: the N of \"Speaker N\" in get_transcription's text, or speakerNumber in its segments; it stays the same after naming.",
|
|
479
|
+
"examples": [
|
|
480
|
+
2
|
|
481
|
+
],
|
|
482
|
+
"minimum": 1,
|
|
483
|
+
"type": "integer"
|
|
484
|
+
},
|
|
485
|
+
"speakerId": {
|
|
486
|
+
"description": "Instead of transcriptionId and speaker: a saved speaker's id (id in list_speakers, or speakerId from save_speaker), changed everywhere they are recognised.",
|
|
487
|
+
"examples": [
|
|
488
|
+
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
489
|
+
],
|
|
490
|
+
"format": "uuid",
|
|
491
|
+
"type": "string"
|
|
492
|
+
},
|
|
493
|
+
"transcriptionId": {
|
|
494
|
+
"description": "With speaker: the transcription the speaker is in (from create_transcription, list_transcriptions, search_transcriptions or start_broadcast). Leave out with speakerId.",
|
|
495
|
+
"examples": [
|
|
496
|
+
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
497
|
+
],
|
|
498
|
+
"format": "uuid",
|
|
499
|
+
"type": "string"
|
|
500
|
+
}
|
|
501
|
+
},
|
|
502
|
+
"type": "object"
|
|
503
|
+
},
|
|
504
|
+
"annotations": {
|
|
505
|
+
"title": "Update a speaker",
|
|
506
|
+
"readOnlyHint": false,
|
|
507
|
+
"destructiveHint": false,
|
|
508
|
+
"idempotentHint": true,
|
|
509
|
+
"openWorldHint": false
|
|
510
|
+
}
|
|
511
|
+
},
|
|
532
512
|
{
|
|
533
513
|
"name": "delete_speaker",
|
|
534
514
|
"title": "Delete a saved speaker",
|
|
535
|
-
"description": "Delete a saved speaker for good: their name, fingerprint, stored sample and any cloned voice are erased at once and cannot be restored.\nUse: the person should no longer be recognised, or asked for their voice to be removed. A name inside one transcription:
|
|
515
|
+
"description": "Delete a saved speaker for good: their name, fingerprint, stored sample and any cloned voice are erased at once and cannot be restored.\nUse: the person should no longer be recognised, or asked for their voice to be removed. A name inside one transcription: update_speaker.\nArguments: speakerId is a saved speaker's id, the id field of list_speakers or the speakerId save_speaker returned; not a speaker number (\"Speaker 2\") and not a transcriptionId. To unname someone in one transcription instead: update_speaker with an empty name.\nBehaviour: free. Transcripts keep every line they spoke; lines labelled with their saved name go back to \"Speaker N\".\nReturns: speakerId, deleted true.\nErrors: not_found for an unknown or already deleted id; forbidden for a read-only key.",
|
|
536
516
|
"inputSchema": {
|
|
537
517
|
"additionalProperties": false,
|
|
538
518
|
"properties": {
|
|
@@ -660,10 +640,85 @@
|
|
|
660
640
|
"openWorldHint": true
|
|
661
641
|
}
|
|
662
642
|
},
|
|
643
|
+
{
|
|
644
|
+
"name": "update_broadcast",
|
|
645
|
+
"title": "Update a broadcast",
|
|
646
|
+
"description": "Change a broadcast's settings: its name, who can listen (public, or private with a password), the language spoken, and whether translations are read aloud, in stock or the speakers' own cloned voices.\nUse: before or between runs; settings apply from the next start_broadcast, the name and access at once. Only the fields given change; give at least one. Starting it: start_broadcast; ending it: stop_broadcast; removing it: delete_broadcast.\nArguments: broadcastId (required) from list_broadcasts or start_broadcast. name 1 to 120 characters. access private needs a password (at least 4 characters) the first time; a new password replaces the old one. spokenLanguage a code like en, or auto. spokenTranslations true or false. clonedVoices true needs consent true, recording that everyone speaking agreed to be cloned.\nBehaviour: free; the link stays the same. Changing access or the password signs out every current listener, who must open the link again. Calling again with the same fields changes nothing else.\nReturns: the broadcast (broadcastId, name, link, access, onAir, listenersNow) with spokenLanguage, spokenTranslations and clonedVoices.\nErrors: invalid_options for nothing to change, a bad name, a short password, private without a password, an unknown language, or clonedVoices true without consent; invalid_state for a broadcast in the trash; not_found for an unknown broadcastId; forbidden for a read-only key.",
|
|
647
|
+
"inputSchema": {
|
|
648
|
+
"additionalProperties": false,
|
|
649
|
+
"properties": {
|
|
650
|
+
"access": {
|
|
651
|
+
"description": "public: anyone with the link can follow it; private: listeners must enter password. Changing it signs out every current listener.",
|
|
652
|
+
"enum": [
|
|
653
|
+
"public",
|
|
654
|
+
"private"
|
|
655
|
+
]
|
|
656
|
+
},
|
|
657
|
+
"broadcastId": {
|
|
658
|
+
"description": "A broadcastId from list_broadcasts or start_broadcast.",
|
|
659
|
+
"examples": [
|
|
660
|
+
"0190f0e0-7c1a-7d3e-9b2a-5f4e3d2c1b0a"
|
|
661
|
+
],
|
|
662
|
+
"format": "uuid",
|
|
663
|
+
"type": "string"
|
|
664
|
+
},
|
|
665
|
+
"clonedVoices": {
|
|
666
|
+
"description": "With spokenTranslations: true reads each speaker's translated words in that speaker's own cloned voice (needs consent true); false: stock voices.",
|
|
667
|
+
"type": "boolean"
|
|
668
|
+
},
|
|
669
|
+
"consent": {
|
|
670
|
+
"default": false,
|
|
671
|
+
"description": "Must be true with clonedVoices true: confirms everyone speaking on it agreed to have their voice cloned. Recorded with your account and the time. Not needed to turn cloned voices off.",
|
|
672
|
+
"type": "boolean"
|
|
673
|
+
},
|
|
674
|
+
"name": {
|
|
675
|
+
"description": "New name shown to listeners; trimmed, 1 to 120 characters.",
|
|
676
|
+
"examples": [
|
|
677
|
+
"Sunday service"
|
|
678
|
+
],
|
|
679
|
+
"maxLength": 120,
|
|
680
|
+
"minLength": 1,
|
|
681
|
+
"type": "string"
|
|
682
|
+
},
|
|
683
|
+
"password": {
|
|
684
|
+
"description": "The listeners' password, at least 4 characters: required to make a broadcast private the first time, and setting a new one signs out every current listener. Ignored with access public.",
|
|
685
|
+
"examples": [
|
|
686
|
+
"choir2026"
|
|
687
|
+
],
|
|
688
|
+
"maxLength": 200,
|
|
689
|
+
"minLength": 4,
|
|
690
|
+
"type": "string"
|
|
691
|
+
},
|
|
692
|
+
"spokenLanguage": {
|
|
693
|
+
"description": "Code of the language spoken (e.g. en, es), or auto to detect it. Applies from the next start_broadcast.",
|
|
694
|
+
"examples": [
|
|
695
|
+
"en",
|
|
696
|
+
"auto"
|
|
697
|
+
],
|
|
698
|
+
"type": "string"
|
|
699
|
+
},
|
|
700
|
+
"spokenTranslations": {
|
|
701
|
+
"description": "true: each translation is also read aloud, so listeners can hear it in their language; false: text only. Applies from the next start_broadcast. Listening stays free.",
|
|
702
|
+
"type": "boolean"
|
|
703
|
+
}
|
|
704
|
+
},
|
|
705
|
+
"required": [
|
|
706
|
+
"broadcastId"
|
|
707
|
+
],
|
|
708
|
+
"type": "object"
|
|
709
|
+
},
|
|
710
|
+
"annotations": {
|
|
711
|
+
"title": "Update a broadcast",
|
|
712
|
+
"readOnlyHint": false,
|
|
713
|
+
"destructiveHint": false,
|
|
714
|
+
"idempotentHint": true,
|
|
715
|
+
"openWorldHint": false
|
|
716
|
+
}
|
|
717
|
+
},
|
|
663
718
|
{
|
|
664
719
|
"name": "stop_broadcast",
|
|
665
720
|
"title": "Stop a broadcast",
|
|
666
|
-
"description": "Stop the live transcription a broadcast is running, which also stops its billing.\nUse: when the event is over. To remove the transcript afterwards: delete_transcription; the broadcast and its link: delete_broadcast.\nArguments: broadcastId of a broadcast on air: from start_broadcast's result, or list_broadcasts (onAir true); not the transcriptionId. A second call is refused with invalid_state and changes nothing, so a retry is safe.\nBehaviour: free. The transcript is kept (get_transcription
|
|
721
|
+
"description": "Stop the live transcription a broadcast is running, which also stops its billing.\nUse: when the event is over. To remove the transcript afterwards: delete_transcription; the broadcast and its link: delete_broadcast.\nArguments: broadcastId of a broadcast on air: from start_broadcast's result, or list_broadcasts (onAir true); not the transcriptionId. A second call is refused with invalid_state and changes nothing, so a retry is safe.\nBehaviour: free. The transcript is kept (get_transcription reads it, and its summary, once state is done); the broadcastId and its link stay, ready for the next start_broadcast.\nReturns: the stopped transcription's status (transcriptionId, name, kind live, state finishing then done, startedAt, endedAt, durationS) with broadcastId and link.\nErrors: invalid_state when not on air, not_found for an unknown broadcastId, forbidden for a key without the recordings:publish scope.",
|
|
667
722
|
"inputSchema": {
|
|
668
723
|
"additionalProperties": false,
|
|
669
724
|
"properties": {
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@vocalstack/mcp",
|
|
3
|
-
"version": "1.
|
|
3
|
+
"version": "1.5.0",
|
|
4
4
|
"description": "VocalStack's MCP server: transcribe, translate, search and summarize audio and video, and run live broadcasts, from any MCP client, on VocalStack's public API.",
|
|
5
5
|
"license": "MIT",
|
|
6
6
|
"type": "module",
|