@dalmia/calibrate-mcp 0.0.41 → 0.0.42
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/bin/mcp-server.js +592 -138
- package/bin/mcp-server.js.map +14 -14
- package/esm/landing-page.js +1 -1
- package/esm/lib/config.d.ts +2 -2
- package/esm/lib/config.js +2 -2
- package/esm/mcp-server/mcp-server.js +1 -1
- package/esm/mcp-server/server.js +1 -1
- package/esm/models/annotationtaskresponse.js +1 -1
- package/esm/models/annotationtaskresponse.js.map +1 -1
- package/esm/models/evaluatorcreate.d.ts +4 -0
- package/esm/models/evaluatorcreate.d.ts.map +1 -1
- package/esm/models/evaluatorcreate.js +5 -2
- package/esm/models/evaluatorcreate.js.map +1 -1
- package/esm/models/evaluatordetailresponsecompact.d.ts +5 -0
- package/esm/models/evaluatordetailresponsecompact.d.ts.map +1 -1
- package/esm/models/evaluatordetailresponsecompact.js +6 -2
- package/esm/models/evaluatordetailresponsecompact.js.map +1 -1
- package/esm/models/listevaluatorsevaluatorsgetop.d.ts +2 -0
- package/esm/models/listevaluatorsevaluatorsgetop.d.ts.map +1 -1
- package/esm/models/listevaluatorsevaluatorsgetop.js +2 -0
- package/esm/models/listevaluatorsevaluatorsgetop.js.map +1 -1
- package/esm/models/routersevaluatorsevaluatorresponse.d.ts +5 -0
- package/esm/models/routersevaluatorsevaluatorresponse.d.ts.map +1 -1
- package/esm/models/routersevaluatorsevaluatorresponse.js +6 -2
- package/esm/models/routersevaluatorsevaluatorresponse.js.map +1 -1
- package/package.json +1 -1
- package/src/landing-page.ts +1 -1
- package/src/lib/config.ts +2 -2
- package/src/mcp-server/mcp-server.ts +1 -1
- package/src/mcp-server/server.ts +1 -1
- package/src/models/annotationtaskresponse.ts +1 -1
- package/src/models/evaluatorcreate.ts +6 -2
- package/src/models/evaluatordetailresponsecompact.ts +10 -2
- package/src/models/listevaluatorsevaluatorsgetop.ts +2 -0
- package/src/models/routersevaluatorsevaluatorresponse.ts +10 -2
package/esm/landing-page.js
CHANGED
|
@@ -915,7 +915,7 @@ http_headers = { "api-key-auth" = "YOUR_API_KEY_AUTH" }`;
|
|
|
915
915
|
<h1>Instructions</h1>
|
|
916
916
|
<p>One-click installation for Claude Desktop users</p>
|
|
917
917
|
<div class="instruction-item">
|
|
918
|
-
<a href="https://github.com/dalmia/calibrate-mcp/releases/download/v0.0.
|
|
918
|
+
<a href="https://github.com/dalmia/calibrate-mcp/releases/download/v0.0.42/mcp-server.mcpb" download="mcp-server.mcpb" class="action-button header-action" style="display: inline-flex; margin-bottom: 16px;">
|
|
919
919
|
📥 Download MCP Bundle
|
|
920
920
|
</a>
|
|
921
921
|
</div>
|
package/esm/lib/config.d.ts
CHANGED
|
@@ -35,8 +35,8 @@ export declare function serverURLFromOptions(options: SDKOptions): URL | null;
|
|
|
35
35
|
export declare const SDK_METADATA: {
|
|
36
36
|
readonly language: "typescript";
|
|
37
37
|
readonly openapiDocVersion: "0.1.0";
|
|
38
|
-
readonly sdkVersion: "0.0.
|
|
38
|
+
readonly sdkVersion: "0.0.42";
|
|
39
39
|
readonly genVersion: "2.915.1";
|
|
40
|
-
readonly userAgent: "speakeasy-sdk/mcp-typescript 0.0.
|
|
40
|
+
readonly userAgent: "speakeasy-sdk/mcp-typescript 0.0.42 2.915.1 0.1.0 @dalmia/calibrate-mcp";
|
|
41
41
|
};
|
|
42
42
|
//# sourceMappingURL=config.d.ts.map
|
package/esm/lib/config.js
CHANGED
|
@@ -27,8 +27,8 @@ export function serverURLFromOptions(options) {
|
|
|
27
27
|
export const SDK_METADATA = {
|
|
28
28
|
language: "typescript",
|
|
29
29
|
openapiDocVersion: "0.1.0",
|
|
30
|
-
sdkVersion: "0.0.
|
|
30
|
+
sdkVersion: "0.0.42",
|
|
31
31
|
genVersion: "2.915.1",
|
|
32
|
-
userAgent: "speakeasy-sdk/mcp-typescript 0.0.
|
|
32
|
+
userAgent: "speakeasy-sdk/mcp-typescript 0.0.42 2.915.1 0.1.0 @dalmia/calibrate-mcp",
|
|
33
33
|
};
|
|
34
34
|
//# sourceMappingURL=config.js.map
|
package/esm/mcp-server/server.js
CHANGED
|
@@ -49,7 +49,7 @@ import { tool$tracesCreate } from "./tools/tracesCreate.js";
|
|
|
49
49
|
export function createMCPServer(deps) {
|
|
50
50
|
const server = new McpServer({
|
|
51
51
|
name: "CalibrateMcp",
|
|
52
|
-
version: "0.0.
|
|
52
|
+
version: "0.0.42",
|
|
53
53
|
});
|
|
54
54
|
const getClient = deps.getSDK || (() => new CalibrateMcpCore({
|
|
55
55
|
security: deps.security,
|
|
@@ -31,7 +31,7 @@ export const AnnotationTaskResponse$zodSchema = z.object({
|
|
|
31
31
|
evaluators: z.array(z.record(z.string(), z.any())).optional().describe("The evaluators linked to this task, in display order. Fetching one task by ID enriches each with its live version's rubric. Listing tasks returns lightweight metadata (uuid, name, type, position) only"),
|
|
32
32
|
has_agreement: z.boolean().default(false).describe("Whether the task has at least one comparable human-vs-human or human-vs-evaluator pair, computed over all time"),
|
|
33
33
|
item_count: z.int().nullable().optional().describe("Number of items in the task"),
|
|
34
|
-
items: z.array(z.record(z.string(), z.any())).optional().describe("The task's items, each with its agreement stats. You get these when you fetch one task by ID, not when you list tasks"),
|
|
34
|
+
items: z.array(z.record(z.string(), z.any())).optional().describe("The task's items, each with its agreement stats and an `is_tool_call` flag marking the rows a person labels on the tool call rather than a text reply. You get these when you fetch one task by ID, not when you list tasks"),
|
|
35
35
|
jobs: z.array(z.record(z.string(), z.any())).optional().describe("The task's labelling jobs. You get these when you fetch one task by ID, not when you list tasks"),
|
|
36
36
|
name: z.string().describe("Name of the task"),
|
|
37
37
|
type: AnnotationTaskResponseType$zodSchema.describe("Task type. Determines the shape of each item's payload.\n- `stt`: judge a transcript on its own\n- `llm`: judge one response with its conversation history\n- `llm-general`: judge a standalone `input -> output` pair\n- `conversation`: judge a full conversation"),
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"annotationtaskresponse.js","sourceRoot":"","sources":["../../src/models/annotationtaskresponse.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAGzB;;;;;;;;GAQG;AACH,MAAM,CAAC,MAAM,0BAA0B,GAAG;IACxC,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,UAAU,EAAE,aAAa;IACzB,YAAY,EAAE,cAAc;IAC5B,GAAG,EAAE,KAAK;CACF,CAAC;AAcX,MAAM,CAAC,MAAM,oCAAoC,GAAG,CAAC,CAAC,IAAI,CAAC;IACzD,KAAK;IACL,KAAK;IACL,aAAa;IACb,cAAc;IACd,KAAK;CACN,CAAC,CAAC,QAAQ,CACT,qQAAqQ,CACtQ,CAAC;AAgBF,MAAM,CAAC,MAAM,gCAAgC,GAEzC,CAAC,CAAC,MAAM,CAAC;IACX,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,0CAA0C,CAAC;IAC3E,WAAW,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACpD,wBAAwB,CACzB;IACD,UAAU,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACpE,0MAA0M,CAC3M;IACD,aAAa,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,QAAQ,CAChD,gHAAgH,CACjH;IACD,UAAU,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAChD,6BAA6B,CAC9B;IACD,KAAK,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAC/D,
|
|
1
|
+
{"version":3,"file":"annotationtaskresponse.js","sourceRoot":"","sources":["../../src/models/annotationtaskresponse.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAGzB;;;;;;;;GAQG;AACH,MAAM,CAAC,MAAM,0BAA0B,GAAG;IACxC,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,UAAU,EAAE,aAAa;IACzB,YAAY,EAAE,cAAc;IAC5B,GAAG,EAAE,KAAK;CACF,CAAC;AAcX,MAAM,CAAC,MAAM,oCAAoC,GAAG,CAAC,CAAC,IAAI,CAAC;IACzD,KAAK;IACL,KAAK;IACL,aAAa;IACb,cAAc;IACd,KAAK;CACN,CAAC,CAAC,QAAQ,CACT,qQAAqQ,CACtQ,CAAC;AAgBF,MAAM,CAAC,MAAM,gCAAgC,GAEzC,CAAC,CAAC,MAAM,CAAC;IACX,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,0CAA0C,CAAC;IAC3E,WAAW,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACpD,wBAAwB,CACzB;IACD,UAAU,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACpE,0MAA0M,CAC3M;IACD,aAAa,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,QAAQ,CAChD,gHAAgH,CACjH;IACD,UAAU,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAChD,6BAA6B,CAC9B;IACD,KAAK,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAC/D,6NAA6N,CAC9N;IACD,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAC9D,iGAAiG,CAClG;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,kBAAkB,CAAC;IAC7C,IAAI,EAAE,oCAAoC,CAAC,QAAQ,CACjD,qQAAqQ,CACtQ;IACD,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAC7B,+CAA+C,CAChD;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,wBAAwB,CAAC;CACpD,CAAC,CAAC"}
|
|
@@ -11,6 +11,7 @@ import { EvaluatorVersionCreate } from "./evaluatorversioncreate.js";
|
|
|
11
11
|
* - `llm`: a reply with its conversation history
|
|
12
12
|
* - `llm-general`: a standalone input and output pair
|
|
13
13
|
* - `conversation`: a full conversation
|
|
14
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
14
15
|
*/
|
|
15
16
|
export declare const EvaluatorCreateEvaluatorType: {
|
|
16
17
|
readonly Tts: "tts";
|
|
@@ -18,6 +19,7 @@ export declare const EvaluatorCreateEvaluatorType: {
|
|
|
18
19
|
readonly Llm: "llm";
|
|
19
20
|
readonly LlmGeneral: "llm-general";
|
|
20
21
|
readonly Conversation: "conversation";
|
|
22
|
+
readonly ToolCall: "tool-call";
|
|
21
23
|
};
|
|
22
24
|
/**
|
|
23
25
|
* What the evaluator judges:
|
|
@@ -29,6 +31,7 @@ export declare const EvaluatorCreateEvaluatorType: {
|
|
|
29
31
|
* - `llm`: a reply with its conversation history
|
|
30
32
|
* - `llm-general`: a standalone input and output pair
|
|
31
33
|
* - `conversation`: a full conversation
|
|
34
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
32
35
|
*/
|
|
33
36
|
export type EvaluatorCreateEvaluatorType = ClosedEnum<typeof EvaluatorCreateEvaluatorType>;
|
|
34
37
|
export declare const EvaluatorCreateEvaluatorType$zodSchema: z.ZodEnum<{
|
|
@@ -37,6 +40,7 @@ export declare const EvaluatorCreateEvaluatorType$zodSchema: z.ZodEnum<{
|
|
|
37
40
|
stt: "stt";
|
|
38
41
|
llm: "llm";
|
|
39
42
|
"llm-general": "llm-general";
|
|
43
|
+
"tool-call": "tool-call";
|
|
40
44
|
}>;
|
|
41
45
|
/**
|
|
42
46
|
* The modality the judge reads:
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"evaluatorcreate.d.ts","sourceRoot":"","sources":["../../src/models/evaluatorcreate.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,sBAAsB,EAEvB,MAAM,6BAA6B,CAAC;AAErC
|
|
1
|
+
{"version":3,"file":"evaluatorcreate.d.ts","sourceRoot":"","sources":["../../src/models/evaluatorcreate.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,sBAAsB,EAEvB,MAAM,6BAA6B,CAAC;AAErC;;;;;;;;;;;GAWG;AACH,eAAO,MAAM,4BAA4B;;;;;;;CAO/B,CAAC;AACX;;;;;;;;;;;GAWG;AACH,MAAM,MAAM,4BAA4B,GAAG,UAAU,CACnD,OAAO,4BAA4B,CACpC,CAAC;AAEF,eAAO,MAAM,sCAAsC;;;;;;;EASlD,CAAC;AAEF;;;;;;;GAOG;AACH,eAAO,MAAM,uBAAuB;;;CAG1B,CAAC;AACX;;;;;;;GAOG;AACH,MAAM,MAAM,uBAAuB,GAAG,UAAU,CAC9C,OAAO,uBAAuB,CAC/B,CAAC;AAEF,eAAO,MAAM,iCAAiC;;;EAGuB,CAAC;AAEtE;;;;;;;GAOG;AACH,eAAO,MAAM,yBAAyB;;;CAG5B,CAAC;AACX;;;;;;;GAOG;AACH,MAAM,MAAM,yBAAyB,GAAG,UAAU,CAChD,OAAO,yBAAyB,CACjC,CAAC;AAEF,eAAO,MAAM,mCAAmC;;;EAK/C,CAAC;AAEF,MAAM,MAAM,eAAe,GAAG;IAC5B,IAAI,EAAE,MAAM,CAAC;IACb,WAAW,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IACxC,cAAc,CAAC,EAAE,4BAA4B,GAAG,SAAS,CAAC;IAC1D,SAAS,CAAC,EAAE,uBAAuB,GAAG,SAAS,CAAC;IAChD,WAAW,CAAC,EAAE,yBAAyB,GAAG,SAAS,CAAC;IACpD,OAAO,EAAE,sBAAsB,CAAC;CACjC,CAAC;AAEF,eAAO,MAAM,yBAAyB,EAAE,CAAC,CAAC,OAAO,CAAC,eAAe,CAkB/D,CAAC"}
|
|
@@ -13,6 +13,7 @@ import { EvaluatorVersionCreate$zodSchema, } from "./evaluatorversioncreate.js";
|
|
|
13
13
|
* - `llm`: a reply with its conversation history
|
|
14
14
|
* - `llm-general`: a standalone input and output pair
|
|
15
15
|
* - `conversation`: a full conversation
|
|
16
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
16
17
|
*/
|
|
17
18
|
export const EvaluatorCreateEvaluatorType = {
|
|
18
19
|
Tts: "tts",
|
|
@@ -20,6 +21,7 @@ export const EvaluatorCreateEvaluatorType = {
|
|
|
20
21
|
Llm: "llm",
|
|
21
22
|
LlmGeneral: "llm-general",
|
|
22
23
|
Conversation: "conversation",
|
|
24
|
+
ToolCall: "tool-call",
|
|
23
25
|
};
|
|
24
26
|
export const EvaluatorCreateEvaluatorType$zodSchema = z.enum([
|
|
25
27
|
"tts",
|
|
@@ -27,7 +29,8 @@ export const EvaluatorCreateEvaluatorType$zodSchema = z.enum([
|
|
|
27
29
|
"llm",
|
|
28
30
|
"llm-general",
|
|
29
31
|
"conversation",
|
|
30
|
-
|
|
32
|
+
"tool-call",
|
|
33
|
+
]).describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n");
|
|
31
34
|
/**
|
|
32
35
|
* The modality the judge reads:
|
|
33
36
|
*
|
|
@@ -64,7 +67,7 @@ export const EvaluatorCreate$zodSchema = z.object({
|
|
|
64
67
|
data_type: EvaluatorCreateDataType$zodSchema.default("text").describe("The modality the judge reads:\n\n- `text`\n- `audio`\n"),
|
|
65
68
|
description: z.string().nullable().optional().describe("Description. Omit to leave blank"),
|
|
66
69
|
evaluator_type: EvaluatorCreateEvaluatorType$zodSchema.default("llm")
|
|
67
|
-
.describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n"),
|
|
70
|
+
.describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n"),
|
|
68
71
|
name: z.string().describe("Evaluator name, unique within your workspace"),
|
|
69
72
|
output_type: EvaluatorCreateOutputType$zodSchema.default("binary").describe("How the evaluator scores:\n\n- `binary`: pass or fail\n- `rating`: a numeric score, using the scale in `output_config`\n"),
|
|
70
73
|
version: EvaluatorVersionCreate$zodSchema.describe("One version of an evaluator: its judge prompt, model, variables, and rubric."),
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"evaluatorcreate.js","sourceRoot":"","sources":["../../src/models/evaluatorcreate.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,gCAAgC,GACjC,MAAM,6BAA6B,CAAC;AAErC
|
|
1
|
+
{"version":3,"file":"evaluatorcreate.js","sourceRoot":"","sources":["../../src/models/evaluatorcreate.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,gCAAgC,GACjC,MAAM,6BAA6B,CAAC;AAErC;;;;;;;;;;;GAWG;AACH,MAAM,CAAC,MAAM,4BAA4B,GAAG;IAC1C,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,UAAU,EAAE,aAAa;IACzB,YAAY,EAAE,cAAc;IAC5B,QAAQ,EAAE,WAAW;CACb,CAAC;AAiBX,MAAM,CAAC,MAAM,sCAAsC,GAAG,CAAC,CAAC,IAAI,CAAC;IAC3D,KAAK;IACL,KAAK;IACL,KAAK;IACL,aAAa;IACb,cAAc;IACd,WAAW;CACZ,CAAC,CAAC,QAAQ,CACT,uSAAuS,CACxS,CAAC;AAEF;;;;;;;GAOG;AACH,MAAM,CAAC,MAAM,uBAAuB,GAAG;IACrC,IAAI,EAAE,MAAM;IACZ,KAAK,EAAE,OAAO;CACN,CAAC;AAaX,MAAM,CAAC,MAAM,iCAAiC,GAAG,CAAC,CAAC,IAAI,CAAC;IACtD,MAAM;IACN,OAAO;CACR,CAAC,CAAC,QAAQ,CAAC,wDAAwD,CAAC,CAAC;AAEtE;;;;;;;GAOG;AACH,MAAM,CAAC,MAAM,yBAAyB,GAAG;IACvC,MAAM,EAAE,QAAQ;IAChB,MAAM,EAAE,QAAQ;CACR,CAAC;AAaX,MAAM,CAAC,MAAM,mCAAmC,GAAG,CAAC,CAAC,IAAI,CAAC;IACxD,QAAQ;IACR,QAAQ;CACT,CAAC,CAAC,QAAQ,CACT,0HAA0H,CAC3H,CAAC;AAWF,MAAM,CAAC,MAAM,yBAAyB,GAA+B,CAAC,CAAC,MAAM,CAAC;IAC5E,SAAS,EAAE,iCAAiC,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,QAAQ,CACnE,wDAAwD,CACzD;IACD,WAAW,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACpD,kCAAkC,CACnC;IACD,cAAc,EAAE,sCAAsC,CAAC,OAAO,CAAC,KAAK,CAAC;SAClE,QAAQ,CACP,uSAAuS,CACxS;IACH,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,8CAA8C,CAAC;IACzE,WAAW,EAAE,mCAAmC,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,QAAQ,CACzE,0HAA0H,CAC3H;IACD,OAAO,EAAE,gCAAgC,CAAC,QAAQ,CAChD,8EAA8E,CAC/E;CACF,CAAC,CAAC"}
|
|
@@ -11,6 +11,7 @@ import { EvaluatorVersionCompact } from "./evaluatorversioncompact.js";
|
|
|
11
11
|
* - `llm`: a reply with its conversation history
|
|
12
12
|
* - `llm-general`: a standalone input and output pair
|
|
13
13
|
* - `conversation`: a full conversation
|
|
14
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
14
15
|
*/
|
|
15
16
|
export declare const EvaluatorDetailResponseCompactEvaluatorType: {
|
|
16
17
|
readonly Tts: "tts";
|
|
@@ -18,6 +19,7 @@ export declare const EvaluatorDetailResponseCompactEvaluatorType: {
|
|
|
18
19
|
readonly Llm: "llm";
|
|
19
20
|
readonly LlmGeneral: "llm-general";
|
|
20
21
|
readonly Conversation: "conversation";
|
|
22
|
+
readonly ToolCall: "tool-call";
|
|
21
23
|
};
|
|
22
24
|
/**
|
|
23
25
|
* What the evaluator judges:
|
|
@@ -29,6 +31,7 @@ export declare const EvaluatorDetailResponseCompactEvaluatorType: {
|
|
|
29
31
|
* - `llm`: a reply with its conversation history
|
|
30
32
|
* - `llm-general`: a standalone input and output pair
|
|
31
33
|
* - `conversation`: a full conversation
|
|
34
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
32
35
|
*/
|
|
33
36
|
export type EvaluatorDetailResponseCompactEvaluatorType = ClosedEnum<typeof EvaluatorDetailResponseCompactEvaluatorType>;
|
|
34
37
|
export declare const EvaluatorDetailResponseCompactEvaluatorType$zodSchema: z.ZodEnum<{
|
|
@@ -37,6 +40,7 @@ export declare const EvaluatorDetailResponseCompactEvaluatorType$zodSchema: z.Zo
|
|
|
37
40
|
stt: "stt";
|
|
38
41
|
llm: "llm";
|
|
39
42
|
"llm-general": "llm-general";
|
|
43
|
+
"tool-call": "tool-call";
|
|
40
44
|
}>;
|
|
41
45
|
/**
|
|
42
46
|
* The modality the judge reads:
|
|
@@ -96,6 +100,7 @@ export type EvaluatorDetailResponseCompact = {
|
|
|
96
100
|
data_type: EvaluatorDetailResponseCompactDataType;
|
|
97
101
|
output_type: EvaluatorDetailResponseCompactOutputType;
|
|
98
102
|
is_default: boolean;
|
|
103
|
+
is_protected: boolean;
|
|
99
104
|
slug?: string | null | undefined;
|
|
100
105
|
source_default_slug?: string | null | undefined;
|
|
101
106
|
live_version_id?: string | null | undefined;
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"evaluatordetailresponsecompact.d.ts","sourceRoot":"","sources":["../../src/models/evaluatordetailresponsecompact.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,uBAAuB,EAExB,MAAM,8BAA8B,CAAC;AAEtC
|
|
1
|
+
{"version":3,"file":"evaluatordetailresponsecompact.d.ts","sourceRoot":"","sources":["../../src/models/evaluatordetailresponsecompact.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,uBAAuB,EAExB,MAAM,8BAA8B,CAAC;AAEtC;;;;;;;;;;;GAWG;AACH,eAAO,MAAM,2CAA2C;;;;;;;CAO9C,CAAC;AACX;;;;;;;;;;;GAWG;AACH,MAAM,MAAM,2CAA2C,GAAG,UAAU,CAClE,OAAO,2CAA2C,CACnD,CAAC;AAEF,eAAO,MAAM,qDAAqD;;;;;;;EASjE,CAAC;AAEF;;;;;;;GAOG;AACH,eAAO,MAAM,sCAAsC;;;CAGzC,CAAC;AACX;;;;;;;GAOG;AACH,MAAM,MAAM,sCAAsC,GAAG,UAAU,CAC7D,OAAO,sCAAsC,CAC9C,CAAC;AAEF,eAAO,MAAM,gDAAgD;;;EAGQ,CAAC;AAEtE;;;;;;;GAOG;AACH,eAAO,MAAM,wCAAwC;;;CAG3C,CAAC;AACX;;;;;;;GAOG;AACH,MAAM,MAAM,wCAAwC,GAAG,UAAU,CAC/D,OAAO,wCAAwC,CAChD,CAAC;AAEF,eAAO,MAAM,kDAAkD;;;EAK9D,CAAC;AAEF,MAAM,MAAM,8BAA8B,GAAG;IAC3C,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;IACb,WAAW,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IACxC,cAAc,EAAE,2CAA2C,CAAC;IAC5D,SAAS,EAAE,sCAAsC,CAAC;IAClD,WAAW,EAAE,wCAAwC,CAAC;IACtD,UAAU,EAAE,OAAO,CAAC;IACpB,YAAY,EAAE,OAAO,CAAC;IACtB,IAAI,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IACjC,mBAAmB,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IAChD,eAAe,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IAC5C,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,QAAQ,EAAE,KAAK,CAAC,uBAAuB,CAAC,CAAC;CAC1C,CAAC;AAEF,eAAO,MAAM,wCAAwC,EAAE,CAAC,CAAC,OAAO,CAC9D,8BAA8B,CAyC9B,CAAC"}
|
|
@@ -13,6 +13,7 @@ import { EvaluatorVersionCompact$zodSchema, } from "./evaluatorversioncompact.js
|
|
|
13
13
|
* - `llm`: a reply with its conversation history
|
|
14
14
|
* - `llm-general`: a standalone input and output pair
|
|
15
15
|
* - `conversation`: a full conversation
|
|
16
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
16
17
|
*/
|
|
17
18
|
export const EvaluatorDetailResponseCompactEvaluatorType = {
|
|
18
19
|
Tts: "tts",
|
|
@@ -20,6 +21,7 @@ export const EvaluatorDetailResponseCompactEvaluatorType = {
|
|
|
20
21
|
Llm: "llm",
|
|
21
22
|
LlmGeneral: "llm-general",
|
|
22
23
|
Conversation: "conversation",
|
|
24
|
+
ToolCall: "tool-call",
|
|
23
25
|
};
|
|
24
26
|
export const EvaluatorDetailResponseCompactEvaluatorType$zodSchema = z.enum([
|
|
25
27
|
"tts",
|
|
@@ -27,7 +29,8 @@ export const EvaluatorDetailResponseCompactEvaluatorType$zodSchema = z.enum([
|
|
|
27
29
|
"llm",
|
|
28
30
|
"llm-general",
|
|
29
31
|
"conversation",
|
|
30
|
-
|
|
32
|
+
"tool-call",
|
|
33
|
+
]).describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n");
|
|
31
34
|
/**
|
|
32
35
|
* The modality the judge reads:
|
|
33
36
|
*
|
|
@@ -65,8 +68,9 @@ export const EvaluatorDetailResponseCompact$zodSchema = z.object({
|
|
|
65
68
|
data_type: EvaluatorDetailResponseCompactDataType$zodSchema.describe("The modality the judge reads:\n\n- `text`\n- `audio`\n"),
|
|
66
69
|
description: z.string().nullable().optional().describe("What the evaluator checks"),
|
|
67
70
|
evaluator_type: EvaluatorDetailResponseCompactEvaluatorType$zodSchema
|
|
68
|
-
.describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n"),
|
|
71
|
+
.describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n"),
|
|
69
72
|
is_default: z.boolean().describe("True when the evaluator is a built-in default or your workspace's editable copy of one. False for an evaluator you created yourself"),
|
|
73
|
+
is_protected: z.boolean().describe("True when the evaluator is locked. A locked evaluator cannot be deleted, and only its name, description and rubric can change"),
|
|
70
74
|
live_version_id: z.string().nullable().optional().describe("ID of the version that is currently live"),
|
|
71
75
|
name: z.string().describe("Evaluator name"),
|
|
72
76
|
output_type: EvaluatorDetailResponseCompactOutputType$zodSchema.describe("How the evaluator scores:\n\n- `binary`: pass or fail\n- `rating`: a numeric score, using the scale in `output_config`\n"),
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"evaluatordetailresponsecompact.js","sourceRoot":"","sources":["../../src/models/evaluatordetailresponsecompact.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,iCAAiC,GAClC,MAAM,8BAA8B,CAAC;AAEtC
|
|
1
|
+
{"version":3,"file":"evaluatordetailresponsecompact.js","sourceRoot":"","sources":["../../src/models/evaluatordetailresponsecompact.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,iCAAiC,GAClC,MAAM,8BAA8B,CAAC;AAEtC;;;;;;;;;;;GAWG;AACH,MAAM,CAAC,MAAM,2CAA2C,GAAG;IACzD,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,UAAU,EAAE,aAAa;IACzB,YAAY,EAAE,cAAc;IAC5B,QAAQ,EAAE,WAAW;CACb,CAAC;AAiBX,MAAM,CAAC,MAAM,qDAAqD,GAAG,CAAC,CAAC,IAAI,CAAC;IAC1E,KAAK;IACL,KAAK;IACL,KAAK;IACL,aAAa;IACb,cAAc;IACd,WAAW;CACZ,CAAC,CAAC,QAAQ,CACT,uSAAuS,CACxS,CAAC;AAEF;;;;;;;GAOG;AACH,MAAM,CAAC,MAAM,sCAAsC,GAAG;IACpD,IAAI,EAAE,MAAM;IACZ,KAAK,EAAE,OAAO;CACN,CAAC;AAaX,MAAM,CAAC,MAAM,gDAAgD,GAAG,CAAC,CAAC,IAAI,CAAC;IACrE,MAAM;IACN,OAAO;CACR,CAAC,CAAC,QAAQ,CAAC,wDAAwD,CAAC,CAAC;AAEtE;;;;;;;GAOG;AACH,MAAM,CAAC,MAAM,wCAAwC,GAAG;IACtD,MAAM,EAAE,QAAQ;IAChB,MAAM,EAAE,QAAQ;CACR,CAAC;AAaX,MAAM,CAAC,MAAM,kDAAkD,GAAG,CAAC,CAAC,IAAI,CAAC;IACvE,QAAQ;IACR,QAAQ;CACT,CAAC,CAAC,QAAQ,CACT,0HAA0H,CAC3H,CAAC;AAmBF,MAAM,CAAC,MAAM,wCAAwC,GAEjD,CAAC,CAAC,MAAM,CAAC;IACX,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAC7B,+CAA+C,CAChD;IACD,SAAS,EAAE,gDAAgD,CAAC,QAAQ,CAClE,wDAAwD,CACzD;IACD,WAAW,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACpD,2BAA2B,CAC5B;IACD,cAAc,EAAE,qDAAqD;SAClE,QAAQ,CACP,uSAAuS,CACxS;IACH,UAAU,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,QAAQ,CAC9B,qIAAqI,CACtI;IACD,YAAY,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,QAAQ,CAChC,+HAA+H,CAChI;IACD,eAAe,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACxD,0CAA0C,CAC3C;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,gBAAgB,CAAC;IAC3C,WAAW,EAAE,kDAAkD,CAAC,QAAQ,CACtE,0HAA0H,CAC3H;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAC7C,8CAA8C,CAC/C;IACD,mBAAmB,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAC5D,qJAAqJ,CACtJ;IACD,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAC7B,oDAAoD,CACrD;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,cAAc,CAAC;IACzC,QAAQ,EAAE,CAAC,CAAC,KAAK,CAAC,iCAAiC,CAAC,CAAC,QAAQ,CAC3D,oCAAoC,CACrC;CACF,CAAC,CAAC"}
|
|
@@ -11,6 +11,7 @@ export declare const ListEvaluatorsEvaluatorsGetEvaluatorType: {
|
|
|
11
11
|
readonly Llm: "llm";
|
|
12
12
|
readonly LlmGeneral: "llm-general";
|
|
13
13
|
readonly Conversation: "conversation";
|
|
14
|
+
readonly ToolCall: "tool-call";
|
|
14
15
|
};
|
|
15
16
|
/**
|
|
16
17
|
* Filter by what the evaluator judges. Omit for all types
|
|
@@ -22,6 +23,7 @@ export declare const ListEvaluatorsEvaluatorsGetEvaluatorType$zodSchema: z.ZodEn
|
|
|
22
23
|
stt: "stt";
|
|
23
24
|
llm: "llm";
|
|
24
25
|
"llm-general": "llm-general";
|
|
26
|
+
"tool-call": "tool-call";
|
|
25
27
|
}>;
|
|
26
28
|
/**
|
|
27
29
|
* Filter by modality. Omit for all
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"listevaluatorsevaluatorsgetop.d.ts","sourceRoot":"","sources":["../../src/models/listevaluatorsevaluatorsgetop.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,mBAAmB,EAEpB,MAAM,0BAA0B,CAAC;AAClC,OAAO,EACL,kCAAkC,EAEnC,MAAM,yCAAyC,CAAC;AAEjD;;GAEG;AACH,eAAO,MAAM,wCAAwC
|
|
1
|
+
{"version":3,"file":"listevaluatorsevaluatorsgetop.d.ts","sourceRoot":"","sources":["../../src/models/listevaluatorsevaluatorsgetop.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,mBAAmB,EAEpB,MAAM,0BAA0B,CAAC;AAClC,OAAO,EACL,kCAAkC,EAEnC,MAAM,yCAAyC,CAAC;AAEjD;;GAEG;AACH,eAAO,MAAM,wCAAwC;;;;;;;CAO3C,CAAC;AACX;;GAEG;AACH,MAAM,MAAM,wCAAwC,GAAG,UAAU,CAC/D,OAAO,wCAAwC,CAChD,CAAC;AAEF,eAAO,MAAM,kDAAkD;;;;;;;EAOO,CAAC;AAEvE;;GAEG;AACH,eAAO,MAAM,mCAAmC;;;CAGtC,CAAC;AACX;;GAEG;AACH,MAAM,MAAM,mCAAmC,GAAG,UAAU,CAC1D,OAAO,mCAAmC,CAC3C,CAAC;AAEF,eAAO,MAAM,6CAA6C;;;EAGX,CAAC;AAEhD,MAAM,MAAM,kCAAkC,GAAG;IAC/C,cAAc,CAAC,EAAE,wCAAwC,GAAG,IAAI,GAAG,SAAS,CAAC;IAC7E,SAAS,CAAC,EAAE,mCAAmC,GAAG,IAAI,GAAG,SAAS,CAAC;IACnE,gBAAgB,CAAC,EAAE,OAAO,GAAG,SAAS,CAAC;IACvC,CAAC,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IAC9B,KAAK,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IAClC,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC5B,OAAO,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;CACrC,CAAC;AAEF,eAAO,MAAM,4CAA4C,EAAE,CAAC,CAAC,OAAO,CAClE,kCAAkC,CAqBlC,CAAC;AAEH,MAAM,MAAM,mCAAmC,GAC3C,kCAAkC,GAClC,mBAAmB,CAAC;AAExB,eAAO,MAAM,6CAA6C,EAAE,CAAC,CAAC,OAAO,CACnE,mCAAmC,CAInC,CAAC"}
|
|
@@ -13,6 +13,7 @@ export const ListEvaluatorsEvaluatorsGetEvaluatorType = {
|
|
|
13
13
|
Llm: "llm",
|
|
14
14
|
LlmGeneral: "llm-general",
|
|
15
15
|
Conversation: "conversation",
|
|
16
|
+
ToolCall: "tool-call",
|
|
16
17
|
};
|
|
17
18
|
export const ListEvaluatorsEvaluatorsGetEvaluatorType$zodSchema = z.enum([
|
|
18
19
|
"tts",
|
|
@@ -20,6 +21,7 @@ export const ListEvaluatorsEvaluatorsGetEvaluatorType$zodSchema = z.enum([
|
|
|
20
21
|
"llm",
|
|
21
22
|
"llm-general",
|
|
22
23
|
"conversation",
|
|
24
|
+
"tool-call",
|
|
23
25
|
]).describe("Filter by what the evaluator judges. Omit for all types");
|
|
24
26
|
/**
|
|
25
27
|
* Filter by modality. Omit for all
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"listevaluatorsevaluatorsgetop.js","sourceRoot":"","sources":["../../src/models/listevaluatorsevaluatorsgetop.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,6BAA6B,GAC9B,MAAM,0BAA0B,CAAC;AAClC,OAAO,EAEL,4CAA4C,GAC7C,MAAM,yCAAyC,CAAC;AAEjD;;GAEG;AACH,MAAM,CAAC,MAAM,wCAAwC,GAAG;IACtD,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,UAAU,EAAE,aAAa;IACzB,YAAY,EAAE,cAAc;
|
|
1
|
+
{"version":3,"file":"listevaluatorsevaluatorsgetop.js","sourceRoot":"","sources":["../../src/models/listevaluatorsevaluatorsgetop.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,6BAA6B,GAC9B,MAAM,0BAA0B,CAAC;AAClC,OAAO,EAEL,4CAA4C,GAC7C,MAAM,yCAAyC,CAAC;AAEjD;;GAEG;AACH,MAAM,CAAC,MAAM,wCAAwC,GAAG;IACtD,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,UAAU,EAAE,aAAa;IACzB,YAAY,EAAE,cAAc;IAC5B,QAAQ,EAAE,WAAW;CACb,CAAC;AAQX,MAAM,CAAC,MAAM,kDAAkD,GAAG,CAAC,CAAC,IAAI,CAAC;IACvE,KAAK;IACL,KAAK;IACL,KAAK;IACL,aAAa;IACb,cAAc;IACd,WAAW;CACZ,CAAC,CAAC,QAAQ,CAAC,yDAAyD,CAAC,CAAC;AAEvE;;GAEG;AACH,MAAM,CAAC,MAAM,mCAAmC,GAAG;IACjD,IAAI,EAAE,MAAM;IACZ,KAAK,EAAE,OAAO;CACN,CAAC;AAQX,MAAM,CAAC,MAAM,6CAA6C,GAAG,CAAC,CAAC,IAAI,CAAC;IAClE,MAAM;IACN,OAAO;CACR,CAAC,CAAC,QAAQ,CAAC,kCAAkC,CAAC,CAAC;AAYhD,MAAM,CAAC,MAAM,4CAA4C,GAErD,CAAC,CAAC,MAAM,CAAC;IACX,SAAS,EAAE,6CAA6C,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE;SAC3E,QAAQ,CAAC,kCAAkC,CAAC;IAC/C,cAAc,EAAE,kDAAkD,CAAC,QAAQ,EAAE;SAC1E,QAAQ,EAAE,CAAC,QAAQ,CAClB,yDAAyD,CAC1D;IACH,gBAAgB,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,QAAQ,CAClD,iJAAiJ,CAClJ;IACD,KAAK,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,QAAQ,CACrB,kEAAkE,CACnE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE;IACvB,MAAM,EAAE,CAAC,CAAC,GAAG,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,QAAQ,CACjC,kDAAkD,CACnD;IACD,CAAC,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CACpB,+DAA+D,CAChE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE;IACvB,OAAO,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE;CAC1C,CAAC,CAAC;AAMH,MAAM,CAAC,MAAM,6CAA6C,GAEtD,CAAC,CAAC,KAAK,CAAC;IACV,4CAA4C;IAC5C,6BAA6B;CAC9B,CAAC,CAAC"}
|
|
@@ -11,6 +11,7 @@ import { EvaluatorLiveVersionSummary } from "./evaluatorliveversionsummary.js";
|
|
|
11
11
|
* - `llm`: a reply with its conversation history
|
|
12
12
|
* - `llm-general`: a standalone input and output pair
|
|
13
13
|
* - `conversation`: a full conversation
|
|
14
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
14
15
|
*/
|
|
15
16
|
export declare const RoutersEvaluatorsEvaluatorResponseEvaluatorType: {
|
|
16
17
|
readonly Tts: "tts";
|
|
@@ -18,6 +19,7 @@ export declare const RoutersEvaluatorsEvaluatorResponseEvaluatorType: {
|
|
|
18
19
|
readonly Llm: "llm";
|
|
19
20
|
readonly LlmGeneral: "llm-general";
|
|
20
21
|
readonly Conversation: "conversation";
|
|
22
|
+
readonly ToolCall: "tool-call";
|
|
21
23
|
};
|
|
22
24
|
/**
|
|
23
25
|
* What the evaluator judges:
|
|
@@ -29,6 +31,7 @@ export declare const RoutersEvaluatorsEvaluatorResponseEvaluatorType: {
|
|
|
29
31
|
* - `llm`: a reply with its conversation history
|
|
30
32
|
* - `llm-general`: a standalone input and output pair
|
|
31
33
|
* - `conversation`: a full conversation
|
|
34
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
32
35
|
*/
|
|
33
36
|
export type RoutersEvaluatorsEvaluatorResponseEvaluatorType = ClosedEnum<typeof RoutersEvaluatorsEvaluatorResponseEvaluatorType>;
|
|
34
37
|
export declare const RoutersEvaluatorsEvaluatorResponseEvaluatorType$zodSchema: z.ZodEnum<{
|
|
@@ -37,6 +40,7 @@ export declare const RoutersEvaluatorsEvaluatorResponseEvaluatorType$zodSchema:
|
|
|
37
40
|
stt: "stt";
|
|
38
41
|
llm: "llm";
|
|
39
42
|
"llm-general": "llm-general";
|
|
43
|
+
"tool-call": "tool-call";
|
|
40
44
|
}>;
|
|
41
45
|
/**
|
|
42
46
|
* The modality the judge reads:
|
|
@@ -96,6 +100,7 @@ export type RoutersEvaluatorsEvaluatorResponse = {
|
|
|
96
100
|
data_type: RoutersEvaluatorsEvaluatorResponseDataType;
|
|
97
101
|
output_type: RoutersEvaluatorsEvaluatorResponseOutputType;
|
|
98
102
|
is_default: boolean;
|
|
103
|
+
is_protected: boolean;
|
|
99
104
|
slug?: string | null | undefined;
|
|
100
105
|
source_default_slug?: string | null | undefined;
|
|
101
106
|
live_version_id?: string | null | undefined;
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"routersevaluatorsevaluatorresponse.d.ts","sourceRoot":"","sources":["../../src/models/routersevaluatorsevaluatorresponse.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,2BAA2B,EAE5B,MAAM,kCAAkC,CAAC;AAE1C
|
|
1
|
+
{"version":3,"file":"routersevaluatorsevaluatorresponse.d.ts","sourceRoot":"","sources":["../../src/models/routersevaluatorsevaluatorresponse.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AACzB,OAAO,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAC/C,OAAO,EACL,2BAA2B,EAE5B,MAAM,kCAAkC,CAAC;AAE1C;;;;;;;;;;;GAWG;AACH,eAAO,MAAM,+CAA+C;;;;;;;CAOlD,CAAC;AACX;;;;;;;;;;;GAWG;AACH,MAAM,MAAM,+CAA+C,GAAG,UAAU,CACtE,OAAO,+CAA+C,CACvD,CAAC;AAEF,eAAO,MAAM,yDAAyD;;;;;;;EAWrE,CAAC;AAEF;;;;;;;GAOG;AACH,eAAO,MAAM,0CAA0C;;;CAG7C,CAAC;AACX;;;;;;;GAOG;AACH,MAAM,MAAM,0CAA0C,GAAG,UAAU,CACjE,OAAO,0CAA0C,CAClD,CAAC;AAEF,eAAO,MAAM,oDAAoD;;;EAGI,CAAC;AAEtE;;;;;;;GAOG;AACH,eAAO,MAAM,4CAA4C;;;CAG/C,CAAC;AACX;;;;;;;GAOG;AACH,MAAM,MAAM,4CAA4C,GAAG,UAAU,CACnE,OAAO,4CAA4C,CACpD,CAAC;AAEF,eAAO,MAAM,sDAAsD;;;EAKlE,CAAC;AAEF,MAAM,MAAM,kCAAkC,GAAG;IAC/C,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;IACb,WAAW,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IACxC,cAAc,EAAE,+CAA+C,CAAC;IAChE,SAAS,EAAE,0CAA0C,CAAC;IACtD,WAAW,EAAE,4CAA4C,CAAC;IAC1D,UAAU,EAAE,OAAO,CAAC;IACpB,YAAY,EAAE,OAAO,CAAC;IACtB,IAAI,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IACjC,mBAAmB,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IAChD,eAAe,CAAC,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,CAAC;IAC5C,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,YAAY,CAAC,EAAE,2BAA2B,GAAG,IAAI,GAAG,SAAS,CAAC;CAC/D,CAAC;AAEF,eAAO,MAAM,4CAA4C,EAAE,CAAC,CAAC,OAAO,CAClE,kCAAkC,CAwClC,CAAC"}
|
|
@@ -13,6 +13,7 @@ import { EvaluatorLiveVersionSummary$zodSchema, } from "./evaluatorliveversionsu
|
|
|
13
13
|
* - `llm`: a reply with its conversation history
|
|
14
14
|
* - `llm-general`: a standalone input and output pair
|
|
15
15
|
* - `conversation`: a full conversation
|
|
16
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
16
17
|
*/
|
|
17
18
|
export const RoutersEvaluatorsEvaluatorResponseEvaluatorType = {
|
|
18
19
|
Tts: "tts",
|
|
@@ -20,6 +21,7 @@ export const RoutersEvaluatorsEvaluatorResponseEvaluatorType = {
|
|
|
20
21
|
Llm: "llm",
|
|
21
22
|
LlmGeneral: "llm-general",
|
|
22
23
|
Conversation: "conversation",
|
|
24
|
+
ToolCall: "tool-call",
|
|
23
25
|
};
|
|
24
26
|
export const RoutersEvaluatorsEvaluatorResponseEvaluatorType$zodSchema = z.enum([
|
|
25
27
|
"tts",
|
|
@@ -27,7 +29,8 @@ export const RoutersEvaluatorsEvaluatorResponseEvaluatorType$zodSchema = z.enum(
|
|
|
27
29
|
"llm",
|
|
28
30
|
"llm-general",
|
|
29
31
|
"conversation",
|
|
30
|
-
|
|
32
|
+
"tool-call",
|
|
33
|
+
]).describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n");
|
|
31
34
|
/**
|
|
32
35
|
* The modality the judge reads:
|
|
33
36
|
*
|
|
@@ -65,8 +68,9 @@ export const RoutersEvaluatorsEvaluatorResponse$zodSchema = z.object({
|
|
|
65
68
|
data_type: RoutersEvaluatorsEvaluatorResponseDataType$zodSchema.describe("The modality the judge reads:\n\n- `text`\n- `audio`\n"),
|
|
66
69
|
description: z.string().nullable().optional().describe("What the evaluator checks"),
|
|
67
70
|
evaluator_type: RoutersEvaluatorsEvaluatorResponseEvaluatorType$zodSchema
|
|
68
|
-
.describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n"),
|
|
71
|
+
.describe("What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n"),
|
|
69
72
|
is_default: z.boolean().describe("True when the evaluator is a built-in default or your workspace's editable copy of one. False for an evaluator you created yourself"),
|
|
73
|
+
is_protected: z.boolean().describe("True when the evaluator is locked. A locked evaluator cannot be deleted, and only its name, description and rubric can change"),
|
|
70
74
|
live_version: EvaluatorLiveVersionSummary$zodSchema.nullable().optional()
|
|
71
75
|
.describe("The version that is currently live"),
|
|
72
76
|
live_version_id: z.string().nullable().optional().describe("ID of the version that is currently live"),
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"routersevaluatorsevaluatorresponse.js","sourceRoot":"","sources":["../../src/models/routersevaluatorsevaluatorresponse.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,qCAAqC,GACtC,MAAM,kCAAkC,CAAC;AAE1C
|
|
1
|
+
{"version":3,"file":"routersevaluatorsevaluatorresponse.js","sourceRoot":"","sources":["../../src/models/routersevaluatorsevaluatorresponse.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,CAAC,MAAM,KAAK,CAAC;AAEzB,OAAO,EAEL,qCAAqC,GACtC,MAAM,kCAAkC,CAAC;AAE1C;;;;;;;;;;;GAWG;AACH,MAAM,CAAC,MAAM,+CAA+C,GAAG;IAC7D,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,GAAG,EAAE,KAAK;IACV,UAAU,EAAE,aAAa;IACzB,YAAY,EAAE,cAAc;IAC5B,QAAQ,EAAE,WAAW;CACb,CAAC;AAiBX,MAAM,CAAC,MAAM,yDAAyD,GAAG,CAAC,CAAC,IAAI,CAC7E;IACE,KAAK;IACL,KAAK;IACL,KAAK;IACL,aAAa;IACb,cAAc;IACd,WAAW;CACZ,CACF,CAAC,QAAQ,CACR,uSAAuS,CACxS,CAAC;AAEF;;;;;;;GAOG;AACH,MAAM,CAAC,MAAM,0CAA0C,GAAG;IACxD,IAAI,EAAE,MAAM;IACZ,KAAK,EAAE,OAAO;CACN,CAAC;AAaX,MAAM,CAAC,MAAM,oDAAoD,GAAG,CAAC,CAAC,IAAI,CAAC;IACzE,MAAM;IACN,OAAO;CACR,CAAC,CAAC,QAAQ,CAAC,wDAAwD,CAAC,CAAC;AAEtE;;;;;;;GAOG;AACH,MAAM,CAAC,MAAM,4CAA4C,GAAG;IAC1D,MAAM,EAAE,QAAQ;IAChB,MAAM,EAAE,QAAQ;CACR,CAAC;AAaX,MAAM,CAAC,MAAM,sDAAsD,GAAG,CAAC,CAAC,IAAI,CAAC;IAC3E,QAAQ;IACR,QAAQ;CACT,CAAC,CAAC,QAAQ,CACT,0HAA0H,CAC3H,CAAC;AAmBF,MAAM,CAAC,MAAM,4CAA4C,GAErD,CAAC,CAAC,MAAM,CAAC;IACX,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAC7B,+CAA+C,CAChD;IACD,SAAS,EAAE,oDAAoD,CAAC,QAAQ,CACtE,wDAAwD,CACzD;IACD,WAAW,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACpD,2BAA2B,CAC5B;IACD,cAAc,EAAE,yDAAyD;SACtE,QAAQ,CACP,uSAAuS,CACxS;IACH,UAAU,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,QAAQ,CAC9B,qIAAqI,CACtI;IACD,YAAY,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,QAAQ,CAChC,+HAA+H,CAChI;IACD,YAAY,EAAE,qCAAqC,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE;SACtE,QAAQ,CAAC,oCAAoC,CAAC;IACjD,eAAe,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CACxD,0CAA0C,CAC3C;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,gBAAgB,CAAC;IAC3C,WAAW,EAAE,sDAAsD,CAAC,QAAQ,CAC1E,0HAA0H,CAC3H;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAC7C,8CAA8C,CAC/C;IACD,mBAAmB,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,EAAE,CAAC,QAAQ,CAC5D,qJAAqJ,CACtJ;IACD,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAC7B,oDAAoD,CACrD;IACD,IAAI,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,cAAc,CAAC;CAC1C,CAAC,CAAC"}
|
package/package.json
CHANGED
package/src/landing-page.ts
CHANGED
|
@@ -930,7 +930,7 @@ http_headers = { "api-key-auth" = "YOUR_API_KEY_AUTH" }`;
|
|
|
930
930
|
<h1>Instructions</h1>
|
|
931
931
|
<p>One-click installation for Claude Desktop users</p>
|
|
932
932
|
<div class="instruction-item">
|
|
933
|
-
<a href="https://github.com/dalmia/calibrate-mcp/releases/download/v0.0.
|
|
933
|
+
<a href="https://github.com/dalmia/calibrate-mcp/releases/download/v0.0.42/mcp-server.mcpb" download="mcp-server.mcpb" class="action-button header-action" style="display: inline-flex; margin-bottom: 16px;">
|
|
934
934
|
📥 Download MCP Bundle
|
|
935
935
|
</a>
|
|
936
936
|
</div>
|
package/src/lib/config.ts
CHANGED
|
@@ -65,8 +65,8 @@ export function serverURLFromOptions(options: SDKOptions): URL | null {
|
|
|
65
65
|
export const SDK_METADATA = {
|
|
66
66
|
language: "typescript",
|
|
67
67
|
openapiDocVersion: "0.1.0",
|
|
68
|
-
sdkVersion: "0.0.
|
|
68
|
+
sdkVersion: "0.0.42",
|
|
69
69
|
genVersion: "2.915.1",
|
|
70
70
|
userAgent:
|
|
71
|
-
"speakeasy-sdk/mcp-typescript 0.0.
|
|
71
|
+
"speakeasy-sdk/mcp-typescript 0.0.42 2.915.1 0.1.0 @dalmia/calibrate-mcp",
|
|
72
72
|
} as const;
|
package/src/mcp-server/server.ts
CHANGED
|
@@ -75,7 +75,7 @@ export const AnnotationTaskResponse$zodSchema: z.ZodType<
|
|
|
75
75
|
"Number of items in the task",
|
|
76
76
|
),
|
|
77
77
|
items: z.array(z.record(z.string(), z.any())).optional().describe(
|
|
78
|
-
"The task's items, each with its agreement stats. You get these when you fetch one task by ID, not when you list tasks",
|
|
78
|
+
"The task's items, each with its agreement stats and an `is_tool_call` flag marking the rows a person labels on the tool call rather than a text reply. You get these when you fetch one task by ID, not when you list tasks",
|
|
79
79
|
),
|
|
80
80
|
jobs: z.array(z.record(z.string(), z.any())).optional().describe(
|
|
81
81
|
"The task's labelling jobs. You get these when you fetch one task by ID, not when you list tasks",
|
|
@@ -19,6 +19,7 @@ import {
|
|
|
19
19
|
* - `llm`: a reply with its conversation history
|
|
20
20
|
* - `llm-general`: a standalone input and output pair
|
|
21
21
|
* - `conversation`: a full conversation
|
|
22
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
22
23
|
*/
|
|
23
24
|
export const EvaluatorCreateEvaluatorType = {
|
|
24
25
|
Tts: "tts",
|
|
@@ -26,6 +27,7 @@ export const EvaluatorCreateEvaluatorType = {
|
|
|
26
27
|
Llm: "llm",
|
|
27
28
|
LlmGeneral: "llm-general",
|
|
28
29
|
Conversation: "conversation",
|
|
30
|
+
ToolCall: "tool-call",
|
|
29
31
|
} as const;
|
|
30
32
|
/**
|
|
31
33
|
* What the evaluator judges:
|
|
@@ -37,6 +39,7 @@ export const EvaluatorCreateEvaluatorType = {
|
|
|
37
39
|
* - `llm`: a reply with its conversation history
|
|
38
40
|
* - `llm-general`: a standalone input and output pair
|
|
39
41
|
* - `conversation`: a full conversation
|
|
42
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
40
43
|
*/
|
|
41
44
|
export type EvaluatorCreateEvaluatorType = ClosedEnum<
|
|
42
45
|
typeof EvaluatorCreateEvaluatorType
|
|
@@ -48,8 +51,9 @@ export const EvaluatorCreateEvaluatorType$zodSchema = z.enum([
|
|
|
48
51
|
"llm",
|
|
49
52
|
"llm-general",
|
|
50
53
|
"conversation",
|
|
54
|
+
"tool-call",
|
|
51
55
|
]).describe(
|
|
52
|
-
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n",
|
|
56
|
+
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n",
|
|
53
57
|
);
|
|
54
58
|
|
|
55
59
|
/**
|
|
@@ -130,7 +134,7 @@ export const EvaluatorCreate$zodSchema: z.ZodType<EvaluatorCreate> = z.object({
|
|
|
130
134
|
),
|
|
131
135
|
evaluator_type: EvaluatorCreateEvaluatorType$zodSchema.default("llm")
|
|
132
136
|
.describe(
|
|
133
|
-
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n",
|
|
137
|
+
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n",
|
|
134
138
|
),
|
|
135
139
|
name: z.string().describe("Evaluator name, unique within your workspace"),
|
|
136
140
|
output_type: EvaluatorCreateOutputType$zodSchema.default("binary").describe(
|
|
@@ -19,6 +19,7 @@ import {
|
|
|
19
19
|
* - `llm`: a reply with its conversation history
|
|
20
20
|
* - `llm-general`: a standalone input and output pair
|
|
21
21
|
* - `conversation`: a full conversation
|
|
22
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
22
23
|
*/
|
|
23
24
|
export const EvaluatorDetailResponseCompactEvaluatorType = {
|
|
24
25
|
Tts: "tts",
|
|
@@ -26,6 +27,7 @@ export const EvaluatorDetailResponseCompactEvaluatorType = {
|
|
|
26
27
|
Llm: "llm",
|
|
27
28
|
LlmGeneral: "llm-general",
|
|
28
29
|
Conversation: "conversation",
|
|
30
|
+
ToolCall: "tool-call",
|
|
29
31
|
} as const;
|
|
30
32
|
/**
|
|
31
33
|
* What the evaluator judges:
|
|
@@ -37,6 +39,7 @@ export const EvaluatorDetailResponseCompactEvaluatorType = {
|
|
|
37
39
|
* - `llm`: a reply with its conversation history
|
|
38
40
|
* - `llm-general`: a standalone input and output pair
|
|
39
41
|
* - `conversation`: a full conversation
|
|
42
|
+
* - `tool-call`: whether the agent called the right tool, labelled by a person
|
|
40
43
|
*/
|
|
41
44
|
export type EvaluatorDetailResponseCompactEvaluatorType = ClosedEnum<
|
|
42
45
|
typeof EvaluatorDetailResponseCompactEvaluatorType
|
|
@@ -48,8 +51,9 @@ export const EvaluatorDetailResponseCompactEvaluatorType$zodSchema = z.enum([
|
|
|
48
51
|
"llm",
|
|
49
52
|
"llm-general",
|
|
50
53
|
"conversation",
|
|
54
|
+
"tool-call",
|
|
51
55
|
]).describe(
|
|
52
|
-
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n",
|
|
56
|
+
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n",
|
|
53
57
|
);
|
|
54
58
|
|
|
55
59
|
/**
|
|
@@ -120,6 +124,7 @@ export type EvaluatorDetailResponseCompact = {
|
|
|
120
124
|
data_type: EvaluatorDetailResponseCompactDataType;
|
|
121
125
|
output_type: EvaluatorDetailResponseCompactOutputType;
|
|
122
126
|
is_default: boolean;
|
|
127
|
+
is_protected: boolean;
|
|
123
128
|
slug?: string | null | undefined;
|
|
124
129
|
source_default_slug?: string | null | undefined;
|
|
125
130
|
live_version_id?: string | null | undefined;
|
|
@@ -142,11 +147,14 @@ export const EvaluatorDetailResponseCompact$zodSchema: z.ZodType<
|
|
|
142
147
|
),
|
|
143
148
|
evaluator_type: EvaluatorDetailResponseCompactEvaluatorType$zodSchema
|
|
144
149
|
.describe(
|
|
145
|
-
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n",
|
|
150
|
+
"What the evaluator judges:\n\n- `tts`: TTS audio\n- `stt`: one transcript\n- `llm`: a reply with its conversation history\n- `llm-general`: a standalone input and output pair\n- `conversation`: a full conversation\n- `tool-call`: whether the agent called the right tool, labelled by a person\n",
|
|
146
151
|
),
|
|
147
152
|
is_default: z.boolean().describe(
|
|
148
153
|
"True when the evaluator is a built-in default or your workspace's editable copy of one. False for an evaluator you created yourself",
|
|
149
154
|
),
|
|
155
|
+
is_protected: z.boolean().describe(
|
|
156
|
+
"True when the evaluator is locked. A locked evaluator cannot be deleted, and only its name, description and rubric can change",
|
|
157
|
+
),
|
|
150
158
|
live_version_id: z.string().nullable().optional().describe(
|
|
151
159
|
"ID of the version that is currently live",
|
|
152
160
|
),
|
|
@@ -22,6 +22,7 @@ export const ListEvaluatorsEvaluatorsGetEvaluatorType = {
|
|
|
22
22
|
Llm: "llm",
|
|
23
23
|
LlmGeneral: "llm-general",
|
|
24
24
|
Conversation: "conversation",
|
|
25
|
+
ToolCall: "tool-call",
|
|
25
26
|
} as const;
|
|
26
27
|
/**
|
|
27
28
|
* Filter by what the evaluator judges. Omit for all types
|
|
@@ -36,6 +37,7 @@ export const ListEvaluatorsEvaluatorsGetEvaluatorType$zodSchema = z.enum([
|
|
|
36
37
|
"llm",
|
|
37
38
|
"llm-general",
|
|
38
39
|
"conversation",
|
|
40
|
+
"tool-call",
|
|
39
41
|
]).describe("Filter by what the evaluator judges. Omit for all types");
|
|
40
42
|
|
|
41
43
|
/**
|