@openpond/evals 0.3.0 → 0.4.0
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/CONTRACT.md +11 -3
- package/README.md +27 -8
- package/RELEASING.md +13 -46
- package/dist/compatibility.js +2 -2
- package/dist/conformance.js +1 -2
- package/dist/evidence/authoring.js +1 -1
- package/dist/evidence/conformance.js +1 -1
- package/dist/evidence/contracts.js +1 -2
- package/dist/evidence/eligibility.js +1 -1
- package/dist/graders.js +1 -1
- package/dist/harness.js +33 -101
- package/dist/index.js +2 -3
- package/dist/model-improvement-qualification.js +125 -0
- package/dist/review-conformance.js +209 -0
- package/dist/runs.js +1 -9
- package/dist/tasksets.js +2 -15
- package/dist/types/compatibility.d.ts +1 -1
- package/dist/types/compatibility.d.ts.map +1 -1
- package/dist/types/conformance.d.ts +2 -0
- package/dist/types/conformance.d.ts.map +1 -1
- package/dist/types/evidence/conformance.d.ts +20 -20
- package/dist/types/evidence/contracts.d.ts +18 -18
- package/dist/types/harness.d.ts +3 -409
- package/dist/types/harness.d.ts.map +1 -1
- package/dist/types/index.d.ts +2 -3
- package/dist/types/index.d.ts.map +1 -1
- package/dist/types/model-improvement-qualification.d.ts +222 -0
- package/dist/types/model-improvement-qualification.d.ts.map +1 -0
- package/dist/types/review-conformance.d.ts +607 -0
- package/dist/types/review-conformance.d.ts.map +1 -0
- package/dist/types/runs.d.ts +1 -9
- package/dist/types/runs.d.ts.map +1 -1
- package/dist/types/tasksets.d.ts +2 -25
- package/dist/types/tasksets.d.ts.map +1 -1
- package/package.json +14 -10
- package/dist/common.js +0 -67
- package/dist/harness-improvements.js +0 -329
- package/dist/harness-workspaces.js +0 -368
- package/dist/sha256.js +0 -91
- package/dist/types/common.d.ts +0 -51
- package/dist/types/common.d.ts.map +0 -1
- package/dist/types/harness-improvements.d.ts +0 -523
- package/dist/types/harness-improvements.d.ts.map +0 -1
- package/dist/types/harness-workspaces.d.ts +0 -802
- package/dist/types/harness-workspaces.d.ts.map +0 -1
- package/dist/types/sha256.d.ts +0 -2
- package/dist/types/sha256.d.ts.map +0 -1
|
@@ -0,0 +1,209 @@
|
|
|
1
|
+
import { contentHash, createHarnessEvaluationReviewReceipt, } from "@openpond/harness";
|
|
2
|
+
import { createModelImprovementQualificationReceipt } from "./model-improvement-qualification.js";
|
|
3
|
+
const createdAt = "2026-08-08T12:00:00.000Z";
|
|
4
|
+
const harnessRelease = ref("harness-release");
|
|
5
|
+
const sourcePolicy = {
|
|
6
|
+
policy: ref("source-policy"),
|
|
7
|
+
state: "authorized",
|
|
8
|
+
checkedAt: createdAt,
|
|
9
|
+
};
|
|
10
|
+
const evidence = (id, kind) => ({
|
|
11
|
+
evidence: ref(id),
|
|
12
|
+
kind,
|
|
13
|
+
sourceRef: `source-${id}`,
|
|
14
|
+
sourcePolicy,
|
|
15
|
+
occurrenceKey: contentHash(`occurrence-${id}`),
|
|
16
|
+
occurredAt: createdAt,
|
|
17
|
+
});
|
|
18
|
+
const watermark = {
|
|
19
|
+
cursor: contentHash("review-watermark"),
|
|
20
|
+
throughCreatedAt: createdAt,
|
|
21
|
+
};
|
|
22
|
+
const claim = (family, count = 3) => ({
|
|
23
|
+
fingerprint: contentHash(`claim-${family}`),
|
|
24
|
+
recurrenceFamily: family,
|
|
25
|
+
statement: `The ${family} behavior remains unresolved after smaller-layer triage.`,
|
|
26
|
+
independentOccurrences: count,
|
|
27
|
+
unresolvedOccurrences: count,
|
|
28
|
+
});
|
|
29
|
+
const noAction = createHarnessEvaluationReviewReceipt({
|
|
30
|
+
schemaVersion: "openpond.harnessEvaluationReviewReceipt.v1",
|
|
31
|
+
id: "review-no-action",
|
|
32
|
+
ownerScope: { kind: "personal", id: "owner-1" },
|
|
33
|
+
workspaceRef: "workspace-1",
|
|
34
|
+
harnessRelease,
|
|
35
|
+
previousWatermark: null,
|
|
36
|
+
nextWatermark: watermark,
|
|
37
|
+
selectedEvidence: [],
|
|
38
|
+
excludedEvidence: [],
|
|
39
|
+
claim: null,
|
|
40
|
+
classification: "no_action",
|
|
41
|
+
triage: [],
|
|
42
|
+
reason: "The bounded evidence window contains no unresolved reusable claim.",
|
|
43
|
+
nextAuthority: "none",
|
|
44
|
+
maxEstimatedCostUsd: 0,
|
|
45
|
+
tasksetProposal: null,
|
|
46
|
+
evaluation: null,
|
|
47
|
+
trainingQualification: null,
|
|
48
|
+
policyVersion: "harness-review-policy-v1",
|
|
49
|
+
createdAt,
|
|
50
|
+
metadata: {},
|
|
51
|
+
});
|
|
52
|
+
const runtime = createHarnessEvaluationReviewReceipt({
|
|
53
|
+
...withoutHash(noAction),
|
|
54
|
+
id: "review-runtime",
|
|
55
|
+
selectedEvidence: [evidence("runtime-failure", "observation")],
|
|
56
|
+
claim: claim("runtime-transport-failure", 1),
|
|
57
|
+
classification: "runtime",
|
|
58
|
+
triage: [
|
|
59
|
+
{
|
|
60
|
+
layer: "runtime",
|
|
61
|
+
status: "unresolved",
|
|
62
|
+
reason: "The adapter failed before model policy could affect the result.",
|
|
63
|
+
evidenceRefs: [ref("runtime-failure")],
|
|
64
|
+
},
|
|
65
|
+
],
|
|
66
|
+
reason: "A deterministic runtime regression is the smallest correct fix.",
|
|
67
|
+
nextAuthority: "runtime_service",
|
|
68
|
+
});
|
|
69
|
+
const product = createHarnessEvaluationReviewReceipt({
|
|
70
|
+
...withoutHash(noAction),
|
|
71
|
+
id: "review-product",
|
|
72
|
+
selectedEvidence: [evidence("product-routing", "route_decision")],
|
|
73
|
+
claim: claim("product-routing-defect", 1),
|
|
74
|
+
classification: "product",
|
|
75
|
+
triage: [
|
|
76
|
+
{
|
|
77
|
+
layer: "product",
|
|
78
|
+
status: "unresolved",
|
|
79
|
+
reason: "The product selected an unrelated Skill for an ordinary Work turn.",
|
|
80
|
+
evidenceRefs: [ref("product-routing")],
|
|
81
|
+
},
|
|
82
|
+
],
|
|
83
|
+
reason: "Product routing must be corrected before behavioral Evaluation.",
|
|
84
|
+
nextAuthority: "product_team",
|
|
85
|
+
});
|
|
86
|
+
const taskset = createHarnessEvaluationReviewReceipt({
|
|
87
|
+
...withoutHash(noAction),
|
|
88
|
+
id: "review-taskset",
|
|
89
|
+
selectedEvidence: [
|
|
90
|
+
evidence("failure-1", "work_outcome"),
|
|
91
|
+
evidence("failure-2", "work_outcome"),
|
|
92
|
+
evidence("failure-3", "work_outcome"),
|
|
93
|
+
],
|
|
94
|
+
claim: claim("search-budget-allocation"),
|
|
95
|
+
classification: "taskset",
|
|
96
|
+
triage: [
|
|
97
|
+
{
|
|
98
|
+
layer: "harness",
|
|
99
|
+
status: "unresolved",
|
|
100
|
+
reason: "The active research Skill did not resolve three independent failures.",
|
|
101
|
+
evidenceRefs: [ref("failure-1"), ref("failure-2"), ref("failure-3")],
|
|
102
|
+
},
|
|
103
|
+
],
|
|
104
|
+
reason: "The repeated behavioral claim now needs controlled measurement.",
|
|
105
|
+
nextAuthority: "human_review",
|
|
106
|
+
tasksetProposal: ref("taskset-proposal"),
|
|
107
|
+
maxEstimatedCostUsd: 2,
|
|
108
|
+
});
|
|
109
|
+
const blockedRl = createModelImprovementQualificationReceipt({
|
|
110
|
+
schemaVersion: "openpond.modelImprovementQualificationReceipt.v1",
|
|
111
|
+
id: "qualification-rl-blocked",
|
|
112
|
+
review: reference(taskset),
|
|
113
|
+
harnessRelease,
|
|
114
|
+
tasksetRelease: ref("taskset-release"),
|
|
115
|
+
baselineEvaluation: ref("baseline-evaluation"),
|
|
116
|
+
model: modelRef(),
|
|
117
|
+
environmentHash: contentHash("environment"),
|
|
118
|
+
toolContractHash: contentHash("tools"),
|
|
119
|
+
permissionContractHash: contentHash("permissions"),
|
|
120
|
+
policyHash: contentHash("policy"),
|
|
121
|
+
verifierRef: ref("verifier"),
|
|
122
|
+
sourcePolicies: [sourcePolicy],
|
|
123
|
+
trainingEvidenceRefs: [ref("training-evidence")],
|
|
124
|
+
frozenEvaluationEvidenceRefs: [ref("frozen-evidence")],
|
|
125
|
+
privacyApproval: ref("privacy-approval"),
|
|
126
|
+
budgetApproval: ref("budget-approval"),
|
|
127
|
+
maximumCostUsd: 25,
|
|
128
|
+
signal: {
|
|
129
|
+
kind: "scalar_reward",
|
|
130
|
+
strength: "weak",
|
|
131
|
+
calibrated: true,
|
|
132
|
+
confounded: false,
|
|
133
|
+
variance: 0,
|
|
134
|
+
evidenceRefs: [ref("reward-audit")],
|
|
135
|
+
},
|
|
136
|
+
decision: "no_training",
|
|
137
|
+
reasons: ["The observed reward is constant and cannot support RL."],
|
|
138
|
+
createdAt,
|
|
139
|
+
metadata: { blockedMethod: "rl" },
|
|
140
|
+
});
|
|
141
|
+
const qualifiedRl = createModelImprovementQualificationReceipt({
|
|
142
|
+
...withoutHash(blockedRl),
|
|
143
|
+
id: "qualification-rl-qualified",
|
|
144
|
+
signal: {
|
|
145
|
+
kind: "scalar_reward",
|
|
146
|
+
strength: "usable",
|
|
147
|
+
calibrated: true,
|
|
148
|
+
confounded: false,
|
|
149
|
+
variance: 0.18,
|
|
150
|
+
evidenceRefs: [ref("reward-audit")],
|
|
151
|
+
},
|
|
152
|
+
decision: "rl",
|
|
153
|
+
reasons: [
|
|
154
|
+
"The frozen baseline has usable variance and a calibrated sequential reward.",
|
|
155
|
+
],
|
|
156
|
+
});
|
|
157
|
+
const modelImprovement = createHarnessEvaluationReviewReceipt({
|
|
158
|
+
...withoutHash(noAction),
|
|
159
|
+
id: "review-model-improvement",
|
|
160
|
+
selectedEvidence: [
|
|
161
|
+
evidence("baseline-evaluation", "evaluation"),
|
|
162
|
+
evidence("qualification-rl-qualified", "training_qualification"),
|
|
163
|
+
],
|
|
164
|
+
claim: claim("search-budget-allocation"),
|
|
165
|
+
classification: "model_improvement",
|
|
166
|
+
triage: [
|
|
167
|
+
{
|
|
168
|
+
layer: "model",
|
|
169
|
+
status: "unresolved",
|
|
170
|
+
reason: "Harness, runtime, product, retrieval, and tool triage left a qualified sequential policy gap.",
|
|
171
|
+
evidenceRefs: [ref("baseline-evaluation"), reference(qualifiedRl)],
|
|
172
|
+
},
|
|
173
|
+
],
|
|
174
|
+
reason: "The qualified claim may proceed to a separately approved managed-training plan.",
|
|
175
|
+
nextAuthority: "training_system",
|
|
176
|
+
tasksetProposal: ref("taskset-proposal"),
|
|
177
|
+
evaluation: ref("baseline-evaluation"),
|
|
178
|
+
trainingQualification: reference(qualifiedRl),
|
|
179
|
+
maxEstimatedCostUsd: 25,
|
|
180
|
+
});
|
|
181
|
+
export const harnessEvaluationReviewConformance = {
|
|
182
|
+
noAction,
|
|
183
|
+
runtime,
|
|
184
|
+
product,
|
|
185
|
+
taskset,
|
|
186
|
+
blockedRl,
|
|
187
|
+
qualifiedRl,
|
|
188
|
+
modelImprovement,
|
|
189
|
+
};
|
|
190
|
+
function ref(id) {
|
|
191
|
+
return { id, contentHash: contentHash(id) };
|
|
192
|
+
}
|
|
193
|
+
function reference(value) {
|
|
194
|
+
return { id: value.id, contentHash: value.contentHash };
|
|
195
|
+
}
|
|
196
|
+
function withoutHash(value) {
|
|
197
|
+
const { contentHash: _contentHash, ...content } = value;
|
|
198
|
+
return content;
|
|
199
|
+
}
|
|
200
|
+
function modelRef() {
|
|
201
|
+
return {
|
|
202
|
+
provider: "openpond",
|
|
203
|
+
model: "openpond-chat",
|
|
204
|
+
revision: "model-revision-1",
|
|
205
|
+
artifactHash: contentHash("model-artifact"),
|
|
206
|
+
tokenizerRevision: "tokenizer-1",
|
|
207
|
+
chatTemplateHash: contentHash("chat-template"),
|
|
208
|
+
};
|
|
209
|
+
}
|
package/dist/runs.js
CHANGED
|
@@ -1,13 +1,5 @@
|
|
|
1
1
|
import { z } from "zod";
|
|
2
|
-
import { FailureClassSchema, ImmutableArtifactRefSchema, ImmutableReleaseRefSchema, MetadataSchema, ReleaseHashSchema, ReleaseIdSchema, ReleaseTimestampSchema, assertContentHash, contentHash, } from "
|
|
3
|
-
export const ModelRefSchema = z.object({
|
|
4
|
-
provider: ReleaseIdSchema,
|
|
5
|
-
model: ReleaseIdSchema,
|
|
6
|
-
revision: z.string().trim().min(1).max(500).nullable().default(null),
|
|
7
|
-
artifactHash: ReleaseHashSchema.nullable().default(null),
|
|
8
|
-
tokenizerRevision: z.string().trim().min(1).max(500).nullable().default(null),
|
|
9
|
-
chatTemplateHash: ReleaseHashSchema.nullable().default(null),
|
|
10
|
-
}).strict();
|
|
2
|
+
import { FailureClassSchema, ImmutableArtifactRefSchema, ImmutableReleaseRefSchema, MetadataSchema, ModelRefSchema, ReleaseHashSchema, ReleaseIdSchema, ReleaseTimestampSchema, assertContentHash, contentHash, } from "@openpond/harness";
|
|
11
3
|
export const RuntimeTargetBindingSchema = z.object({
|
|
12
4
|
adapterId: ReleaseIdSchema,
|
|
13
5
|
placement: z.enum(["local", "remote", "colocated", "provider_native"]),
|
package/dist/tasksets.js
CHANGED
|
@@ -1,5 +1,5 @@
|
|
|
1
1
|
import { z } from "zod";
|
|
2
|
-
import { ImmutableAssetRefSchema, MetadataSchema, ReleaseHashSchema, ReleaseIdSchema, assertContentHash, contentHash, } from "
|
|
2
|
+
import { CapabilityRequirementSchema, ImmutableAssetRefSchema, MetadataSchema, ReleaseHashSchema, ReleaseIdSchema, ToolDeclarationSchema, assertContentHash, contentHash, } from "@openpond/harness";
|
|
3
3
|
export const TaskSplitSchema = z.enum(["train", "validation", "test", "frozen_eval"]);
|
|
4
4
|
export const PolicyBoundarySchema = z.object({
|
|
5
5
|
policyVisibleFields: z.array(ReleaseIdSchema).max(1_000).default([]),
|
|
@@ -7,20 +7,6 @@ export const PolicyBoundarySchema = z.object({
|
|
|
7
7
|
hiddenGraderRefs: z.array(ReleaseIdSchema).max(1_000).default([]),
|
|
8
8
|
connectedAppScopes: z.array(ReleaseIdSchema).max(100).default([]),
|
|
9
9
|
}).strict();
|
|
10
|
-
export const ToolDeclarationSchema = z.object({
|
|
11
|
-
name: z.string().trim().min(1).max(64).regex(/^[a-zA-Z][a-zA-Z0-9_-]*$/),
|
|
12
|
-
description: z.string().trim().min(1).max(2_000),
|
|
13
|
-
inputSchema: z.record(z.string(), z.unknown()),
|
|
14
|
-
inputSchemaHash: ReleaseHashSchema,
|
|
15
|
-
sideEffect: z.enum(["read", "write"]),
|
|
16
|
-
timeoutMs: z.number().int().positive().max(3_600_000),
|
|
17
|
-
}).strict();
|
|
18
|
-
export const CapabilityRequirementSchema = z.object({
|
|
19
|
-
id: ReleaseIdSchema,
|
|
20
|
-
required: z.boolean(),
|
|
21
|
-
scopes: z.array(z.string().trim().min(1).max(500)).max(100).default([]),
|
|
22
|
-
portability: z.enum(["portable", "host_adapter", "local_only", "hosted_only"]),
|
|
23
|
-
}).strict();
|
|
24
10
|
export const EnvironmentContractSchema = z.object({
|
|
25
11
|
protocolVersion: z.literal("openpond.environment.v1"),
|
|
26
12
|
kind: z.enum(["text", "agent", "work", "custom_program"]),
|
|
@@ -160,3 +146,4 @@ export function trainingPolicyTaskViews(taskset) {
|
|
|
160
146
|
.filter((task) => task.split !== "frozen_eval")
|
|
161
147
|
.map(policyTaskView);
|
|
162
148
|
}
|
|
149
|
+
export { CapabilityRequirementSchema, ToolDeclarationSchema, } from "@openpond/harness";
|
|
@@ -1,4 +1,4 @@
|
|
|
1
|
-
import { type HarnessRelease } from "
|
|
1
|
+
import { type HarnessRelease } from "@openpond/harness";
|
|
2
2
|
import { type HarnessCompatibilityReceipt } from "./runs.js";
|
|
3
3
|
import { type TasksetRelease } from "./tasksets.js";
|
|
4
4
|
export declare function createVerifiedHarnessCompatibilityReceipt(input: {
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"compatibility.d.ts","sourceRoot":"","sources":["../../../../src/compatibility.ts"],"names":[],"mappings":"AACA,OAAO,EAEL,KAAK,cAAc,EACpB,MAAM,
|
|
1
|
+
{"version":3,"file":"compatibility.d.ts","sourceRoot":"","sources":["../../../../src/compatibility.ts"],"names":[],"mappings":"AACA,OAAO,EAEL,KAAK,cAAc,EACpB,MAAM,mBAAmB,CAAC;AAC3B,OAAO,EAEL,KAAK,2BAA2B,EACjC,MAAM,WAAW,CAAC;AACnB,OAAO,EAEL,KAAK,cAAc,EACpB,MAAM,eAAe,CAAC;AAEvB,wBAAgB,yCAAyC,CAAC,KAAK,EAAE;IAC/D,EAAE,EAAE,MAAM,CAAC;IACX,kBAAkB,EAAE,cAAc,CAAC;IACnC,uBAAuB,EAAE,cAAc,CAAC;IACxC,cAAc,EAAE,cAAc,CAAC;IAC/B,QAAQ,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACpC,GAAG,2BAA2B,CA4C9B"}
|
|
@@ -42,6 +42,7 @@ export declare const genericToolConformance: {
|
|
|
42
42
|
id: string;
|
|
43
43
|
required: boolean;
|
|
44
44
|
scopes: string[];
|
|
45
|
+
portability: "portable" | "host_adapter" | "local_only" | "hosted_only";
|
|
45
46
|
}[];
|
|
46
47
|
dependencyLock: {
|
|
47
48
|
id: string;
|
|
@@ -308,6 +309,7 @@ export declare const marketingPortfolioConformance: {
|
|
|
308
309
|
id: string;
|
|
309
310
|
required: boolean;
|
|
310
311
|
scopes: string[];
|
|
312
|
+
portability: "portable" | "host_adapter" | "local_only" | "hosted_only";
|
|
311
313
|
}[];
|
|
312
314
|
dependencyLock: {
|
|
313
315
|
id: string;
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"conformance.d.ts","sourceRoot":"","sources":["../../../../src/conformance.ts"],"names":[],"mappings":"
|
|
1
|
+
{"version":3,"file":"conformance.d.ts","sourceRoot":"","sources":["../../../../src/conformance.ts"],"names":[],"mappings":"AAaA,eAAO,MAAM,sBAAsB;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;CAEjC,CAAC;AAEH,eAAO,MAAM,6BAA6B;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;CAGxC,CAAC"}
|
|
@@ -9,8 +9,8 @@ export declare const completeWorkProcessTraceFixture: {
|
|
|
9
9
|
sequence: number;
|
|
10
10
|
timestamp: string;
|
|
11
11
|
layer: "agent" | "environment";
|
|
12
|
-
kind: "
|
|
13
|
-
action: "
|
|
12
|
+
kind: "approval" | "validation" | "artifact" | "tool" | "state_transition" | "question" | "cleanup";
|
|
13
|
+
action: "turn_started" | "turn_completed" | "turn_failed" | "turn_cancelled" | "turn_timed_out" | "tool_invoked" | "tool_completed" | "tool_failed" | "validation_completed" | "validation_failed" | "approval_requested" | "approval_resolved" | "question_asked" | "question_answered" | "question_dismissed" | "artifact_created" | "workspace_changed" | "environment_created" | "environment_reset" | "environment_destroyed" | "cleanup_completed" | "cleanup_failed";
|
|
14
14
|
status: "cancelled" | "failed" | "started" | "completed";
|
|
15
15
|
inputHash: string | null;
|
|
16
16
|
outputHash: string | null;
|
|
@@ -23,7 +23,7 @@ export declare const completeWorkProcessTraceFixture: {
|
|
|
23
23
|
id: string;
|
|
24
24
|
}[];
|
|
25
25
|
attributes: {
|
|
26
|
-
toolCategory: "agent" | "
|
|
26
|
+
toolCategory: "agent" | "filesystem" | "source_control" | "command" | "browser" | "connected_app" | "sandbox" | "other" | null;
|
|
27
27
|
validationKind: "test" | "other" | "structural" | "visual" | "user_review" | null;
|
|
28
28
|
transitionState: "timeout" | "cancelled" | "failed" | "completed" | "running" | null;
|
|
29
29
|
interventionOutcome: "requested" | "approved" | "denied" | "answered" | "dismissed" | null;
|
|
@@ -32,7 +32,7 @@ export declare const completeWorkProcessTraceFixture: {
|
|
|
32
32
|
durationMs: number | null;
|
|
33
33
|
cpuTimeMs: number | null;
|
|
34
34
|
memoryPeakBytes: number | null;
|
|
35
|
-
errorClass: "unknown" | "
|
|
35
|
+
errorClass: "unknown" | "timeout" | "cancelled" | "validation" | "policy" | "environment" | "infrastructure" | null;
|
|
36
36
|
};
|
|
37
37
|
}[];
|
|
38
38
|
contentHash: string;
|
|
@@ -41,7 +41,7 @@ export declare const completeWorkEvidenceFixture: {
|
|
|
41
41
|
schemaVersion: "openpond.workEvidenceReceipt.v1";
|
|
42
42
|
id: string;
|
|
43
43
|
source: {
|
|
44
|
-
surface: "
|
|
44
|
+
surface: "desktop" | "hosted";
|
|
45
45
|
experience: "work" | "development";
|
|
46
46
|
opaqueRef: string;
|
|
47
47
|
revisionHash: string;
|
|
@@ -186,8 +186,8 @@ export declare const incompleteWorkProcessTraceFixture: {
|
|
|
186
186
|
sequence: number;
|
|
187
187
|
timestamp: string;
|
|
188
188
|
layer: "agent" | "environment";
|
|
189
|
-
kind: "
|
|
190
|
-
action: "
|
|
189
|
+
kind: "approval" | "validation" | "artifact" | "tool" | "state_transition" | "question" | "cleanup";
|
|
190
|
+
action: "turn_started" | "turn_completed" | "turn_failed" | "turn_cancelled" | "turn_timed_out" | "tool_invoked" | "tool_completed" | "tool_failed" | "validation_completed" | "validation_failed" | "approval_requested" | "approval_resolved" | "question_asked" | "question_answered" | "question_dismissed" | "artifact_created" | "workspace_changed" | "environment_created" | "environment_reset" | "environment_destroyed" | "cleanup_completed" | "cleanup_failed";
|
|
191
191
|
status: "cancelled" | "failed" | "started" | "completed";
|
|
192
192
|
inputHash: string | null;
|
|
193
193
|
outputHash: string | null;
|
|
@@ -200,7 +200,7 @@ export declare const incompleteWorkProcessTraceFixture: {
|
|
|
200
200
|
id: string;
|
|
201
201
|
}[];
|
|
202
202
|
attributes: {
|
|
203
|
-
toolCategory: "agent" | "
|
|
203
|
+
toolCategory: "agent" | "filesystem" | "source_control" | "command" | "browser" | "connected_app" | "sandbox" | "other" | null;
|
|
204
204
|
validationKind: "test" | "other" | "structural" | "visual" | "user_review" | null;
|
|
205
205
|
transitionState: "timeout" | "cancelled" | "failed" | "completed" | "running" | null;
|
|
206
206
|
interventionOutcome: "requested" | "approved" | "denied" | "answered" | "dismissed" | null;
|
|
@@ -209,7 +209,7 @@ export declare const incompleteWorkProcessTraceFixture: {
|
|
|
209
209
|
durationMs: number | null;
|
|
210
210
|
cpuTimeMs: number | null;
|
|
211
211
|
memoryPeakBytes: number | null;
|
|
212
|
-
errorClass: "unknown" | "
|
|
212
|
+
errorClass: "unknown" | "timeout" | "cancelled" | "validation" | "policy" | "environment" | "infrastructure" | null;
|
|
213
213
|
};
|
|
214
214
|
}[];
|
|
215
215
|
contentHash: string;
|
|
@@ -273,7 +273,7 @@ export declare const revokedWorkEvidenceEligibilityFixture: {
|
|
|
273
273
|
export declare const invalidRawEvidenceFixture: {
|
|
274
274
|
source: {
|
|
275
275
|
opaqueRef: string;
|
|
276
|
-
surface: "
|
|
276
|
+
surface: "desktop" | "hosted";
|
|
277
277
|
experience: "work" | "development";
|
|
278
278
|
revisionHash: string;
|
|
279
279
|
};
|
|
@@ -392,8 +392,8 @@ export declare const workEvidenceConformance: {
|
|
|
392
392
|
sequence: number;
|
|
393
393
|
timestamp: string;
|
|
394
394
|
layer: "agent" | "environment";
|
|
395
|
-
kind: "
|
|
396
|
-
action: "
|
|
395
|
+
kind: "approval" | "validation" | "artifact" | "tool" | "state_transition" | "question" | "cleanup";
|
|
396
|
+
action: "turn_started" | "turn_completed" | "turn_failed" | "turn_cancelled" | "turn_timed_out" | "tool_invoked" | "tool_completed" | "tool_failed" | "validation_completed" | "validation_failed" | "approval_requested" | "approval_resolved" | "question_asked" | "question_answered" | "question_dismissed" | "artifact_created" | "workspace_changed" | "environment_created" | "environment_reset" | "environment_destroyed" | "cleanup_completed" | "cleanup_failed";
|
|
397
397
|
status: "cancelled" | "failed" | "started" | "completed";
|
|
398
398
|
inputHash: string | null;
|
|
399
399
|
outputHash: string | null;
|
|
@@ -406,7 +406,7 @@ export declare const workEvidenceConformance: {
|
|
|
406
406
|
id: string;
|
|
407
407
|
}[];
|
|
408
408
|
attributes: {
|
|
409
|
-
toolCategory: "agent" | "
|
|
409
|
+
toolCategory: "agent" | "filesystem" | "source_control" | "command" | "browser" | "connected_app" | "sandbox" | "other" | null;
|
|
410
410
|
validationKind: "test" | "other" | "structural" | "visual" | "user_review" | null;
|
|
411
411
|
transitionState: "timeout" | "cancelled" | "failed" | "completed" | "running" | null;
|
|
412
412
|
interventionOutcome: "requested" | "approved" | "denied" | "answered" | "dismissed" | null;
|
|
@@ -415,7 +415,7 @@ export declare const workEvidenceConformance: {
|
|
|
415
415
|
durationMs: number | null;
|
|
416
416
|
cpuTimeMs: number | null;
|
|
417
417
|
memoryPeakBytes: number | null;
|
|
418
|
-
errorClass: "unknown" | "
|
|
418
|
+
errorClass: "unknown" | "timeout" | "cancelled" | "validation" | "policy" | "environment" | "infrastructure" | null;
|
|
419
419
|
};
|
|
420
420
|
}[];
|
|
421
421
|
contentHash: string;
|
|
@@ -424,7 +424,7 @@ export declare const workEvidenceConformance: {
|
|
|
424
424
|
schemaVersion: "openpond.workEvidenceReceipt.v1";
|
|
425
425
|
id: string;
|
|
426
426
|
source: {
|
|
427
|
-
surface: "
|
|
427
|
+
surface: "desktop" | "hosted";
|
|
428
428
|
experience: "work" | "development";
|
|
429
429
|
opaqueRef: string;
|
|
430
430
|
revisionHash: string;
|
|
@@ -625,8 +625,8 @@ export declare const workEvidenceConformance: {
|
|
|
625
625
|
sequence: number;
|
|
626
626
|
timestamp: string;
|
|
627
627
|
layer: "agent" | "environment";
|
|
628
|
-
kind: "
|
|
629
|
-
action: "
|
|
628
|
+
kind: "approval" | "validation" | "artifact" | "tool" | "state_transition" | "question" | "cleanup";
|
|
629
|
+
action: "turn_started" | "turn_completed" | "turn_failed" | "turn_cancelled" | "turn_timed_out" | "tool_invoked" | "tool_completed" | "tool_failed" | "validation_completed" | "validation_failed" | "approval_requested" | "approval_resolved" | "question_asked" | "question_answered" | "question_dismissed" | "artifact_created" | "workspace_changed" | "environment_created" | "environment_reset" | "environment_destroyed" | "cleanup_completed" | "cleanup_failed";
|
|
630
630
|
status: "cancelled" | "failed" | "started" | "completed";
|
|
631
631
|
inputHash: string | null;
|
|
632
632
|
outputHash: string | null;
|
|
@@ -639,7 +639,7 @@ export declare const workEvidenceConformance: {
|
|
|
639
639
|
id: string;
|
|
640
640
|
}[];
|
|
641
641
|
attributes: {
|
|
642
|
-
toolCategory: "agent" | "
|
|
642
|
+
toolCategory: "agent" | "filesystem" | "source_control" | "command" | "browser" | "connected_app" | "sandbox" | "other" | null;
|
|
643
643
|
validationKind: "test" | "other" | "structural" | "visual" | "user_review" | null;
|
|
644
644
|
transitionState: "timeout" | "cancelled" | "failed" | "completed" | "running" | null;
|
|
645
645
|
interventionOutcome: "requested" | "approved" | "denied" | "answered" | "dismissed" | null;
|
|
@@ -648,7 +648,7 @@ export declare const workEvidenceConformance: {
|
|
|
648
648
|
durationMs: number | null;
|
|
649
649
|
cpuTimeMs: number | null;
|
|
650
650
|
memoryPeakBytes: number | null;
|
|
651
|
-
errorClass: "unknown" | "
|
|
651
|
+
errorClass: "unknown" | "timeout" | "cancelled" | "validation" | "policy" | "environment" | "infrastructure" | null;
|
|
652
652
|
};
|
|
653
653
|
}[];
|
|
654
654
|
contentHash: string;
|
|
@@ -656,7 +656,7 @@ export declare const workEvidenceConformance: {
|
|
|
656
656
|
readonly invalidRawEvidence: {
|
|
657
657
|
source: {
|
|
658
658
|
opaqueRef: string;
|
|
659
|
-
surface: "
|
|
659
|
+
surface: "desktop" | "hosted";
|
|
660
660
|
experience: "work" | "development";
|
|
661
661
|
revisionHash: string;
|
|
662
662
|
};
|
|
@@ -13,9 +13,9 @@ export declare const EvidenceArtifactRefSchema: z.ZodObject<{
|
|
|
13
13
|
id: z.ZodString;
|
|
14
14
|
}, z.core.$strict>;
|
|
15
15
|
export declare const WorkProcessStepKindSchema: z.ZodEnum<{
|
|
16
|
+
approval: "approval";
|
|
16
17
|
validation: "validation";
|
|
17
18
|
artifact: "artifact";
|
|
18
|
-
approval: "approval";
|
|
19
19
|
tool: "tool";
|
|
20
20
|
state_transition: "state_transition";
|
|
21
21
|
question: "question";
|
|
@@ -26,8 +26,8 @@ export declare const WorkProcessLayerSchema: z.ZodEnum<{
|
|
|
26
26
|
environment: "environment";
|
|
27
27
|
}>;
|
|
28
28
|
export declare const WorkProcessActionSchema: z.ZodEnum<{
|
|
29
|
-
turn_completed: "turn_completed";
|
|
30
29
|
turn_started: "turn_started";
|
|
30
|
+
turn_completed: "turn_completed";
|
|
31
31
|
turn_failed: "turn_failed";
|
|
32
32
|
turn_cancelled: "turn_cancelled";
|
|
33
33
|
turn_timed_out: "turn_timed_out";
|
|
@@ -57,11 +57,11 @@ export declare const WorkProcessStepStatusSchema: z.ZodEnum<{
|
|
|
57
57
|
}>;
|
|
58
58
|
export declare const WorkToolCategorySchema: z.ZodEnum<{
|
|
59
59
|
agent: "agent";
|
|
60
|
-
connected_app: "connected_app";
|
|
61
60
|
filesystem: "filesystem";
|
|
62
61
|
source_control: "source_control";
|
|
63
62
|
command: "command";
|
|
64
63
|
browser: "browser";
|
|
64
|
+
connected_app: "connected_app";
|
|
65
65
|
sandbox: "sandbox";
|
|
66
66
|
other: "other";
|
|
67
67
|
}>;
|
|
@@ -88,10 +88,10 @@ export declare const WorkInterventionOutcomeSchema: z.ZodEnum<{
|
|
|
88
88
|
}>;
|
|
89
89
|
export declare const WorkProcessErrorClassSchema: z.ZodEnum<{
|
|
90
90
|
unknown: "unknown";
|
|
91
|
-
policy: "policy";
|
|
92
91
|
timeout: "timeout";
|
|
93
92
|
cancelled: "cancelled";
|
|
94
93
|
validation: "validation";
|
|
94
|
+
policy: "policy";
|
|
95
95
|
environment: "environment";
|
|
96
96
|
infrastructure: "infrastructure";
|
|
97
97
|
}>;
|
|
@@ -111,17 +111,17 @@ export declare const WorkProcessStepSchema: z.ZodObject<{
|
|
|
111
111
|
environment: "environment";
|
|
112
112
|
}>;
|
|
113
113
|
kind: z.ZodEnum<{
|
|
114
|
+
approval: "approval";
|
|
114
115
|
validation: "validation";
|
|
115
116
|
artifact: "artifact";
|
|
116
|
-
approval: "approval";
|
|
117
117
|
tool: "tool";
|
|
118
118
|
state_transition: "state_transition";
|
|
119
119
|
question: "question";
|
|
120
120
|
cleanup: "cleanup";
|
|
121
121
|
}>;
|
|
122
122
|
action: z.ZodEnum<{
|
|
123
|
-
turn_completed: "turn_completed";
|
|
124
123
|
turn_started: "turn_started";
|
|
124
|
+
turn_completed: "turn_completed";
|
|
125
125
|
turn_failed: "turn_failed";
|
|
126
126
|
turn_cancelled: "turn_cancelled";
|
|
127
127
|
turn_timed_out: "turn_timed_out";
|
|
@@ -162,11 +162,11 @@ export declare const WorkProcessStepSchema: z.ZodObject<{
|
|
|
162
162
|
attributes: z.ZodObject<{
|
|
163
163
|
toolCategory: z.ZodNullable<z.ZodEnum<{
|
|
164
164
|
agent: "agent";
|
|
165
|
-
connected_app: "connected_app";
|
|
166
165
|
filesystem: "filesystem";
|
|
167
166
|
source_control: "source_control";
|
|
168
167
|
command: "command";
|
|
169
168
|
browser: "browser";
|
|
169
|
+
connected_app: "connected_app";
|
|
170
170
|
sandbox: "sandbox";
|
|
171
171
|
other: "other";
|
|
172
172
|
}>>;
|
|
@@ -198,10 +198,10 @@ export declare const WorkProcessStepSchema: z.ZodObject<{
|
|
|
198
198
|
memoryPeakBytes: z.ZodNullable<z.ZodNumber>;
|
|
199
199
|
errorClass: z.ZodNullable<z.ZodEnum<{
|
|
200
200
|
unknown: "unknown";
|
|
201
|
-
policy: "policy";
|
|
202
201
|
timeout: "timeout";
|
|
203
202
|
cancelled: "cancelled";
|
|
204
203
|
validation: "validation";
|
|
204
|
+
policy: "policy";
|
|
205
205
|
environment: "environment";
|
|
206
206
|
infrastructure: "infrastructure";
|
|
207
207
|
}>>;
|
|
@@ -229,17 +229,17 @@ export declare const WorkProcessTraceContentSchema: z.ZodObject<{
|
|
|
229
229
|
environment: "environment";
|
|
230
230
|
}>;
|
|
231
231
|
kind: z.ZodEnum<{
|
|
232
|
+
approval: "approval";
|
|
232
233
|
validation: "validation";
|
|
233
234
|
artifact: "artifact";
|
|
234
|
-
approval: "approval";
|
|
235
235
|
tool: "tool";
|
|
236
236
|
state_transition: "state_transition";
|
|
237
237
|
question: "question";
|
|
238
238
|
cleanup: "cleanup";
|
|
239
239
|
}>;
|
|
240
240
|
action: z.ZodEnum<{
|
|
241
|
-
turn_completed: "turn_completed";
|
|
242
241
|
turn_started: "turn_started";
|
|
242
|
+
turn_completed: "turn_completed";
|
|
243
243
|
turn_failed: "turn_failed";
|
|
244
244
|
turn_cancelled: "turn_cancelled";
|
|
245
245
|
turn_timed_out: "turn_timed_out";
|
|
@@ -280,11 +280,11 @@ export declare const WorkProcessTraceContentSchema: z.ZodObject<{
|
|
|
280
280
|
attributes: z.ZodObject<{
|
|
281
281
|
toolCategory: z.ZodNullable<z.ZodEnum<{
|
|
282
282
|
agent: "agent";
|
|
283
|
-
connected_app: "connected_app";
|
|
284
283
|
filesystem: "filesystem";
|
|
285
284
|
source_control: "source_control";
|
|
286
285
|
command: "command";
|
|
287
286
|
browser: "browser";
|
|
287
|
+
connected_app: "connected_app";
|
|
288
288
|
sandbox: "sandbox";
|
|
289
289
|
other: "other";
|
|
290
290
|
}>>;
|
|
@@ -316,10 +316,10 @@ export declare const WorkProcessTraceContentSchema: z.ZodObject<{
|
|
|
316
316
|
memoryPeakBytes: z.ZodNullable<z.ZodNumber>;
|
|
317
317
|
errorClass: z.ZodNullable<z.ZodEnum<{
|
|
318
318
|
unknown: "unknown";
|
|
319
|
-
policy: "policy";
|
|
320
319
|
timeout: "timeout";
|
|
321
320
|
cancelled: "cancelled";
|
|
322
321
|
validation: "validation";
|
|
322
|
+
policy: "policy";
|
|
323
323
|
environment: "environment";
|
|
324
324
|
infrastructure: "infrastructure";
|
|
325
325
|
}>>;
|
|
@@ -348,17 +348,17 @@ export declare const WorkProcessTraceSchema: z.ZodObject<{
|
|
|
348
348
|
environment: "environment";
|
|
349
349
|
}>;
|
|
350
350
|
kind: z.ZodEnum<{
|
|
351
|
+
approval: "approval";
|
|
351
352
|
validation: "validation";
|
|
352
353
|
artifact: "artifact";
|
|
353
|
-
approval: "approval";
|
|
354
354
|
tool: "tool";
|
|
355
355
|
state_transition: "state_transition";
|
|
356
356
|
question: "question";
|
|
357
357
|
cleanup: "cleanup";
|
|
358
358
|
}>;
|
|
359
359
|
action: z.ZodEnum<{
|
|
360
|
-
turn_completed: "turn_completed";
|
|
361
360
|
turn_started: "turn_started";
|
|
361
|
+
turn_completed: "turn_completed";
|
|
362
362
|
turn_failed: "turn_failed";
|
|
363
363
|
turn_cancelled: "turn_cancelled";
|
|
364
364
|
turn_timed_out: "turn_timed_out";
|
|
@@ -399,11 +399,11 @@ export declare const WorkProcessTraceSchema: z.ZodObject<{
|
|
|
399
399
|
attributes: z.ZodObject<{
|
|
400
400
|
toolCategory: z.ZodNullable<z.ZodEnum<{
|
|
401
401
|
agent: "agent";
|
|
402
|
-
connected_app: "connected_app";
|
|
403
402
|
filesystem: "filesystem";
|
|
404
403
|
source_control: "source_control";
|
|
405
404
|
command: "command";
|
|
406
405
|
browser: "browser";
|
|
406
|
+
connected_app: "connected_app";
|
|
407
407
|
sandbox: "sandbox";
|
|
408
408
|
other: "other";
|
|
409
409
|
}>>;
|
|
@@ -435,10 +435,10 @@ export declare const WorkProcessTraceSchema: z.ZodObject<{
|
|
|
435
435
|
memoryPeakBytes: z.ZodNullable<z.ZodNumber>;
|
|
436
436
|
errorClass: z.ZodNullable<z.ZodEnum<{
|
|
437
437
|
unknown: "unknown";
|
|
438
|
-
policy: "policy";
|
|
439
438
|
timeout: "timeout";
|
|
440
439
|
cancelled: "cancelled";
|
|
441
440
|
validation: "validation";
|
|
441
|
+
policy: "policy";
|
|
442
442
|
environment: "environment";
|
|
443
443
|
infrastructure: "infrastructure";
|
|
444
444
|
}>>;
|
|
@@ -479,8 +479,8 @@ export declare const WorkEvidenceReceiptContentSchema: z.ZodObject<{
|
|
|
479
479
|
id: z.ZodString;
|
|
480
480
|
source: z.ZodObject<{
|
|
481
481
|
surface: z.ZodEnum<{
|
|
482
|
-
hosted: "hosted";
|
|
483
482
|
desktop: "desktop";
|
|
483
|
+
hosted: "hosted";
|
|
484
484
|
}>;
|
|
485
485
|
experience: z.ZodEnum<{
|
|
486
486
|
work: "work";
|
|
@@ -609,8 +609,8 @@ export declare const WorkEvidenceReceiptSchema: z.ZodObject<{
|
|
|
609
609
|
id: z.ZodString;
|
|
610
610
|
source: z.ZodObject<{
|
|
611
611
|
surface: z.ZodEnum<{
|
|
612
|
-
hosted: "hosted";
|
|
613
612
|
desktop: "desktop";
|
|
613
|
+
hosted: "hosted";
|
|
614
614
|
}>;
|
|
615
615
|
experience: z.ZodEnum<{
|
|
616
616
|
work: "work";
|