agent-hitch 0.1.1 → 0.2.1
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 +235 -132
- package/README.zh-CN.md +335 -0
- package/dist/bin/hitch.js +9 -0
- package/dist/bin/hitch.js.map +1 -0
- package/dist/scripts/check-architecture.js +159 -0
- package/dist/scripts/check-architecture.js.map +1 -0
- package/dist/scripts/check-release.js +25 -0
- package/dist/scripts/check-release.js.map +1 -0
- package/dist/scripts/check-syntax.js +32 -0
- package/dist/scripts/check-syntax.js.map +1 -0
- package/dist/src/adapters/catalog.js +55 -0
- package/dist/src/adapters/catalog.js.map +1 -0
- package/dist/src/adapters/contract.js +2 -0
- package/dist/src/adapters/contract.js.map +1 -0
- package/dist/src/adapters/discovery.js +22 -0
- package/dist/src/adapters/discovery.js.map +1 -0
- package/dist/src/adapters/index.js +3 -0
- package/dist/src/adapters/index.js.map +1 -0
- package/dist/src/adapters/providers/claude.js +69 -0
- package/dist/src/adapters/providers/claude.js.map +1 -0
- package/dist/src/adapters/providers/codex.js +74 -0
- package/dist/src/adapters/providers/codex.js.map +1 -0
- package/dist/src/adapters/providers/deepseek.js +51 -0
- package/dist/src/adapters/providers/deepseek.js.map +1 -0
- package/dist/src/adapters/providers/opencode.js +73 -0
- package/dist/src/adapters/providers/opencode.js.map +1 -0
- package/dist/src/adapters/providers/pi.js +82 -0
- package/dist/src/adapters/providers/pi.js.map +1 -0
- package/dist/src/adapters/providers/shared.js +82 -0
- package/dist/src/adapters/providers/shared.js.map +1 -0
- package/dist/src/artifacts/index.js +3 -0
- package/dist/src/artifacts/index.js.map +1 -0
- package/dist/src/artifacts/integrity.js +188 -0
- package/dist/src/artifacts/integrity.js.map +1 -0
- package/dist/src/artifacts/preparer.js +379 -0
- package/dist/src/artifacts/preparer.js.map +1 -0
- package/dist/src/artifacts/store.js +87 -0
- package/dist/src/artifacts/store.js.map +1 -0
- package/dist/src/artifacts/types.js +2 -0
- package/dist/src/artifacts/types.js.map +1 -0
- package/dist/src/backends/contract.js +2 -0
- package/dist/src/backends/contract.js.map +1 -0
- package/dist/src/backends/harbor/backend.js +305 -0
- package/dist/src/backends/harbor/backend.js.map +1 -0
- package/dist/src/backends/harbor/index.js +4 -0
- package/dist/src/backends/harbor/index.js.map +1 -0
- package/dist/src/backends/harbor/local-git-transport.js +447 -0
- package/dist/src/backends/harbor/local-git-transport.js.map +1 -0
- package/dist/src/backends/harbor/tools.js +314 -0
- package/dist/src/backends/harbor/tools.js.map +1 -0
- package/dist/src/backends/index.js +4 -0
- package/dist/src/backends/index.js.map +1 -0
- package/dist/src/cli/arguments.js +123 -0
- package/dist/src/cli/arguments.js.map +1 -0
- package/dist/src/cli/commands/compare.js +55 -0
- package/dist/src/cli/commands/compare.js.map +1 -0
- package/dist/src/cli/commands/daemon.js +106 -0
- package/dist/src/cli/commands/daemon.js.map +1 -0
- package/dist/src/cli/commands/eval.js +126 -0
- package/dist/src/cli/commands/eval.js.map +1 -0
- package/dist/src/cli/commands/feedback.js +84 -0
- package/dist/src/cli/commands/feedback.js.map +1 -0
- package/dist/src/cli/commands/inspect.js +20 -0
- package/dist/src/cli/commands/inspect.js.map +1 -0
- package/dist/src/cli/commands/list.js +17 -0
- package/dist/src/cli/commands/list.js.map +1 -0
- package/dist/src/cli/commands/prepare.js +21 -0
- package/dist/src/cli/commands/prepare.js.map +1 -0
- package/dist/src/cli/commands/resolve.js +17 -0
- package/dist/src/cli/commands/resolve.js.map +1 -0
- package/dist/src/cli/commands/run.js +83 -0
- package/dist/src/cli/commands/run.js.map +1 -0
- package/dist/src/cli/commands/runs.js +45 -0
- package/dist/src/cli/commands/runs.js.map +1 -0
- package/dist/src/cli/commands/trajectory.js +31 -0
- package/dist/src/cli/commands/trajectory.js.map +1 -0
- package/dist/src/cli/commands/workspace.js +44 -0
- package/dist/src/cli/commands/workspace.js.map +1 -0
- package/dist/src/cli/index.js +2 -0
- package/dist/src/cli/index.js.map +1 -0
- package/dist/src/cli/main.js +48 -0
- package/dist/src/cli/main.js.map +1 -0
- package/dist/src/cli/output.js +96 -0
- package/dist/src/cli/output.js.map +1 -0
- package/dist/src/controller-runtime/hash.js +353 -0
- package/dist/src/controller-runtime/hash.js.map +1 -0
- package/dist/src/controller-runtime/index.js +3 -0
- package/dist/src/controller-runtime/index.js.map +1 -0
- package/dist/src/controller-runtime/store.js +318 -0
- package/dist/src/controller-runtime/store.js.map +1 -0
- package/dist/src/daemon/auth.js +108 -0
- package/dist/src/daemon/auth.js.map +1 -0
- package/dist/src/daemon/client.js +76 -0
- package/dist/src/daemon/client.js.map +1 -0
- package/dist/src/daemon/index.js +5 -0
- package/dist/src/daemon/index.js.map +1 -0
- package/dist/src/daemon/launcher.js +24 -0
- package/dist/src/daemon/launcher.js.map +1 -0
- package/dist/src/daemon/scheduler.js +161 -0
- package/dist/src/daemon/scheduler.js.map +1 -0
- package/dist/src/daemon/server.js +316 -0
- package/dist/src/daemon/server.js.map +1 -0
- package/dist/src/domain/artifacts.js +2 -0
- package/dist/src/domain/artifacts.js.map +1 -0
- package/dist/src/domain/controller-runtime.js +2 -0
- package/dist/src/domain/controller-runtime.js.map +1 -0
- package/dist/src/domain/eval-records.js +2 -0
- package/dist/src/domain/eval-records.js.map +1 -0
- package/dist/src/domain/evals.js +2 -0
- package/dist/src/domain/evals.js.map +1 -0
- package/dist/src/domain/feedback.js +2 -0
- package/dist/src/domain/feedback.js.map +1 -0
- package/dist/src/domain/ids.js +10 -0
- package/dist/src/domain/ids.js.map +1 -0
- package/dist/src/domain/index.js +11 -0
- package/dist/src/domain/index.js.map +1 -0
- package/dist/src/domain/revisions.js +2 -0
- package/dist/src/domain/revisions.js.map +1 -0
- package/dist/src/domain/runs.js +3 -0
- package/dist/src/domain/runs.js.map +1 -0
- package/dist/src/domain/trajectories.js +2 -0
- package/dist/src/domain/trajectories.js.map +1 -0
- package/dist/src/domain/validation.js +429 -0
- package/dist/src/domain/validation.js.map +1 -0
- package/dist/src/domain/workspaces.js +2 -0
- package/dist/src/domain/workspaces.js.map +1 -0
- package/dist/src/evals/events.js +67 -0
- package/dist/src/evals/events.js.map +1 -0
- package/dist/src/evals/index.js +5 -0
- package/dist/src/evals/index.js.map +1 -0
- package/dist/src/evals/records.js +55 -0
- package/dist/src/evals/records.js.map +1 -0
- package/dist/src/evals/request.js +109 -0
- package/dist/src/evals/request.js.map +1 -0
- package/dist/src/evals/service.js +227 -0
- package/dist/src/evals/service.js.map +1 -0
- package/dist/src/evals/trial-import.js +394 -0
- package/dist/src/evals/trial-import.js.map +1 -0
- package/dist/src/feedback/index.js +2 -0
- package/dist/src/feedback/index.js.map +1 -0
- package/dist/src/feedback/service.js +314 -0
- package/dist/src/feedback/service.js.map +1 -0
- package/dist/src/foundation/config.js +54 -0
- package/dist/src/foundation/config.js.map +1 -0
- package/dist/src/foundation/errors.js +14 -0
- package/dist/src/foundation/errors.js.map +1 -0
- package/dist/src/foundation/executable.js +135 -0
- package/dist/src/foundation/executable.js.map +1 -0
- package/dist/src/foundation/fs.js +45 -0
- package/dist/src/foundation/fs.js.map +1 -0
- package/dist/src/foundation/hash.js +24 -0
- package/dist/src/foundation/hash.js.map +1 -0
- package/dist/src/foundation/index.js +11 -0
- package/dist/src/foundation/index.js.map +1 -0
- package/dist/src/foundation/line-stream.js +20 -0
- package/dist/src/foundation/line-stream.js.map +1 -0
- package/dist/src/foundation/locks.js +114 -0
- package/dist/src/foundation/locks.js.map +1 -0
- package/dist/src/foundation/package-root.js +41 -0
- package/dist/src/foundation/package-root.js.map +1 -0
- package/dist/src/foundation/process.js +50 -0
- package/dist/src/foundation/process.js.map +1 -0
- package/dist/src/revisions/index.js +4 -0
- package/dist/src/revisions/index.js.map +1 -0
- package/dist/src/revisions/reference.js +97 -0
- package/dist/src/revisions/reference.js.map +1 -0
- package/dist/src/revisions/resolver.js +25 -0
- package/dist/src/revisions/resolver.js.map +1 -0
- package/dist/src/revisions/sources/git.js +180 -0
- package/dist/src/revisions/sources/git.js.map +1 -0
- package/dist/src/revisions/sources/installed.js +45 -0
- package/dist/src/revisions/sources/installed.js.map +1 -0
- package/dist/src/revisions/sources/npm.js +97 -0
- package/dist/src/revisions/sources/npm.js.map +1 -0
- package/dist/src/runs/compare.js +132 -0
- package/dist/src/runs/compare.js.map +1 -0
- package/dist/src/runs/events.js +79 -0
- package/dist/src/runs/events.js.map +1 -0
- package/dist/src/runs/executor.js +480 -0
- package/dist/src/runs/executor.js.map +1 -0
- package/dist/src/runs/finalizer.js +22 -0
- package/dist/src/runs/finalizer.js.map +1 -0
- package/dist/src/runs/identity.js +35 -0
- package/dist/src/runs/identity.js.map +1 -0
- package/dist/src/runs/index.js +11 -0
- package/dist/src/runs/index.js.map +1 -0
- package/dist/src/runs/manifest.js +70 -0
- package/dist/src/runs/manifest.js.map +1 -0
- package/dist/src/runs/outcome.js +49 -0
- package/dist/src/runs/outcome.js.map +1 -0
- package/dist/src/runs/query.js +104 -0
- package/dist/src/runs/query.js.map +1 -0
- package/dist/src/runs/queued.js +54 -0
- package/dist/src/runs/queued.js.map +1 -0
- package/dist/src/runs/records.js +288 -0
- package/dist/src/runs/records.js.map +1 -0
- package/dist/src/runs/request.js +169 -0
- package/dist/src/runs/request.js.map +1 -0
- package/dist/src/trajectories/contract.js +17 -0
- package/dist/src/trajectories/contract.js.map +1 -0
- package/dist/src/trajectories/format.js +127 -0
- package/dist/src/trajectories/format.js.map +1 -0
- package/dist/src/trajectories/index.js +7 -0
- package/dist/src/trajectories/index.js.map +1 -0
- package/dist/src/trajectories/projector.js +385 -0
- package/dist/src/trajectories/projector.js.map +1 -0
- package/dist/src/trajectories/provider-capture.js +133 -0
- package/dist/src/trajectories/provider-capture.js.map +1 -0
- package/dist/src/trajectories/providers/deepseek.js +179 -0
- package/dist/src/trajectories/providers/deepseek.js.map +1 -0
- package/dist/src/trajectories/store.js +421 -0
- package/dist/src/trajectories/store.js.map +1 -0
- package/dist/src/workspaces/copy.js +142 -0
- package/dist/src/workspaces/copy.js.map +1 -0
- package/dist/src/workspaces/digest.js +47 -0
- package/dist/src/workspaces/digest.js.map +1 -0
- package/dist/src/workspaces/git.js +164 -0
- package/dist/src/workspaces/git.js.map +1 -0
- package/dist/src/workspaces/index.js +6 -0
- package/dist/src/workspaces/index.js.map +1 -0
- package/dist/src/workspaces/lifecycle.js +313 -0
- package/dist/src/workspaces/lifecycle.js.map +1 -0
- package/dist/src/workspaces/planner.js +80 -0
- package/dist/src/workspaces/planner.js.map +1 -0
- package/dist/src/workspaces/store.js +25 -0
- package/dist/src/workspaces/store.js.map +1 -0
- package/dist/src/workspaces/types.js +4 -0
- package/dist/src/workspaces/types.js.map +1 -0
- package/dist/src/workspaces/utils.js +56 -0
- package/dist/src/workspaces/utils.js.map +1 -0
- package/docs/schemas/controller-runtime-manifest.schema.json +57 -0
- package/docs/schemas/controller-runtime-ref.schema.json +16 -0
- package/docs/schemas/eval-request.schema.json +2 -0
- package/docs/schemas/eval-result.schema.json +36 -1
- package/docs/schemas/local-git-transport-manifest.schema.json +34 -0
- package/docs/schemas/message-feedback.schema.json +37 -0
- package/docs/schemas/result.schema.json +9 -0
- package/docs/schemas/run-context.schema.json +39 -0
- package/docs/schemas/run-manifest.schema.json +81 -0
- package/docs/schemas/run-request.schema.json +16 -1
- package/docs/schemas/trajectory-ref.schema.json +73 -0
- package/integrations/harbor/hitch_harbor_agent.py +551 -5
- package/package.json +37 -11
- package/bin/hitch.js +0 -9
- package/src/adapters.js +0 -435
- package/src/artifacts.js +0 -949
- package/src/cli.js +0 -505
- package/src/config.js +0 -51
- package/src/daemon.js +0 -416
- package/src/engine.js +0 -400
- package/src/errors.js +0 -12
- package/src/eval-tools.js +0 -334
- package/src/evals.js +0 -276
- package/src/events.js +0 -69
- package/src/fs.js +0 -43
- package/src/harbor-backend.js +0 -285
- package/src/harness-reference.js +0 -78
- package/src/line-stream.js +0 -17
- package/src/locks.js +0 -33
- package/src/process.js +0 -49
- package/src/registry.js +0 -84
- package/src/scheduler.js +0 -156
- package/src/workspaces.js +0 -893
|
@@ -0,0 +1,109 @@
|
|
|
1
|
+
export const DEFAULT_EVAL_TIMEOUT_MS = 15 * 60 * 1_000;
|
|
2
|
+
export const DEFAULT_EVAL_SETUP_TIMEOUT_MS = 30 * 60 * 1_000;
|
|
3
|
+
const HARBOR_EVAL_BYPASS_ARG = {
|
|
4
|
+
codex: "--dangerously-bypass-approvals-and-sandbox",
|
|
5
|
+
opencode: "--dangerously-skip-permissions",
|
|
6
|
+
};
|
|
7
|
+
export function newEvalId() {
|
|
8
|
+
return `eval_${randomUUID().replaceAll("-", "")}`;
|
|
9
|
+
}
|
|
10
|
+
export async function validateEvalRequest(input) {
|
|
11
|
+
if (!input || typeof input !== "object" || Array.isArray(input))
|
|
12
|
+
throw invalidInput("eval request must be a JSON object");
|
|
13
|
+
const allowed = new Set([
|
|
14
|
+
"schema_version", "backend", "dataset", "harness_ref", "model", "attempts",
|
|
15
|
+
"max_concurrent", "timeout_ms", "setup_timeout_ms", "agent_args", "pass_env",
|
|
16
|
+
]);
|
|
17
|
+
const unexpected = Object.keys(input).find((field) => !allowed.has(field));
|
|
18
|
+
if (unexpected)
|
|
19
|
+
throw invalidInput(`unknown eval request field: ${unexpected}`);
|
|
20
|
+
if (input.schema_version !== undefined && input.schema_version !== SCHEMA_VERSION)
|
|
21
|
+
throw invalidInput(`unsupported schema_version: ${input.schema_version}`);
|
|
22
|
+
if ((input.backend || "harbor") !== "harbor")
|
|
23
|
+
throw invalidInput("only the harbor eval backend is currently supported");
|
|
24
|
+
if (typeof input.dataset !== "string" || !input.dataset.trim())
|
|
25
|
+
throw invalidInput("dataset must be a non-empty string");
|
|
26
|
+
if (typeof input.harness_ref !== "string" || !input.harness_ref.trim())
|
|
27
|
+
throw invalidInput("harness_ref must be a non-empty string");
|
|
28
|
+
const reference = parseHarnessReference(input.harness_ref);
|
|
29
|
+
if (reference.selector.type === "installed") {
|
|
30
|
+
throw invalidInput("eval requires an immutable harness ref: version:<exact> or commit:<sha>");
|
|
31
|
+
}
|
|
32
|
+
assertExactLocalGitEvalReference(reference);
|
|
33
|
+
const benchmark = await resolveBenchmarkReference(input.dataset.trim());
|
|
34
|
+
const attempts = positiveInteger(input.attempts ?? 1, "attempts");
|
|
35
|
+
const maxConcurrent = positiveInteger(input.max_concurrent ?? 4, "max_concurrent");
|
|
36
|
+
const timeout = nonNegativeNumber(input.timeout_ms ?? DEFAULT_EVAL_TIMEOUT_MS, "timeout_ms");
|
|
37
|
+
const setupTimeout = nonNegativeNumber(input.setup_timeout_ms ?? DEFAULT_EVAL_SETUP_TIMEOUT_MS, "setup_timeout_ms");
|
|
38
|
+
if (input.model !== undefined && typeof input.model !== "string")
|
|
39
|
+
throw invalidInput("model must be a string");
|
|
40
|
+
if (input.agent_args !== undefined && (!Array.isArray(input.agent_args) || input.agent_args.some((value) => typeof value !== "string"))) {
|
|
41
|
+
throw invalidInput("agent_args must be an array of strings");
|
|
42
|
+
}
|
|
43
|
+
if (input.pass_env !== undefined && (!Array.isArray(input.pass_env) || input.pass_env.some((value) => typeof value !== "string"))) {
|
|
44
|
+
throw invalidInput("pass_env must be an array of strings");
|
|
45
|
+
}
|
|
46
|
+
const agentArgs = Array.isArray(input.agent_args) ? [...input.agent_args] : [];
|
|
47
|
+
const bypassArg = HARBOR_EVAL_BYPASS_ARG[reference.harness_id];
|
|
48
|
+
if (bypassArg && !agentArgs.includes(bypassArg))
|
|
49
|
+
agentArgs.unshift(bypassArg);
|
|
50
|
+
return {
|
|
51
|
+
schema_version: SCHEMA_VERSION,
|
|
52
|
+
backend: "harbor",
|
|
53
|
+
dataset: String(input.dataset).trim(),
|
|
54
|
+
harness_ref: reference.canonical,
|
|
55
|
+
model: typeof input.model === "string" ? input.model : "",
|
|
56
|
+
attempts,
|
|
57
|
+
max_concurrent: maxConcurrent,
|
|
58
|
+
timeout_ms: timeout,
|
|
59
|
+
setup_timeout_ms: setupTimeout,
|
|
60
|
+
agent_args: agentArgs,
|
|
61
|
+
pass_env: [...new Set(Array.isArray(input.pass_env) ? input.pass_env : [])],
|
|
62
|
+
benchmark_id: benchmark.benchmark_id,
|
|
63
|
+
benchmark_revision: benchmark.benchmark_revision,
|
|
64
|
+
};
|
|
65
|
+
}
|
|
66
|
+
export async function resolveBenchmarkReference(dataset) {
|
|
67
|
+
const raw = dataset.trim();
|
|
68
|
+
const local = path.resolve(raw);
|
|
69
|
+
try {
|
|
70
|
+
if ((await stat(local)).isDirectory()) {
|
|
71
|
+
return {
|
|
72
|
+
benchmark_id: `local:${path.basename(local)}`,
|
|
73
|
+
benchmark_revision: await workspaceDigest(local, { excludedTopLevel: new Set([".git"]) }),
|
|
74
|
+
};
|
|
75
|
+
}
|
|
76
|
+
}
|
|
77
|
+
catch (error) {
|
|
78
|
+
if (error.code !== "ENOENT")
|
|
79
|
+
throw error;
|
|
80
|
+
}
|
|
81
|
+
const separator = raw.lastIndexOf("@");
|
|
82
|
+
if (separator <= 0 || separator === raw.length - 1) {
|
|
83
|
+
throw invalidInput("dataset must include an immutable revision (for example benchmark@1.0)");
|
|
84
|
+
}
|
|
85
|
+
const benchmarkId = raw.slice(0, separator);
|
|
86
|
+
const revision = raw.slice(separator + 1);
|
|
87
|
+
if (revision.toLowerCase() === "latest")
|
|
88
|
+
throw invalidInput("dataset revision cannot be latest");
|
|
89
|
+
return { benchmark_id: benchmarkId, benchmark_revision: revision };
|
|
90
|
+
}
|
|
91
|
+
function positiveInteger(value, name) {
|
|
92
|
+
const parsed = Number(value);
|
|
93
|
+
if (!Number.isInteger(parsed) || parsed <= 0)
|
|
94
|
+
throw invalidInput(`${name} must be a positive integer`);
|
|
95
|
+
return parsed;
|
|
96
|
+
}
|
|
97
|
+
function nonNegativeNumber(value, name) {
|
|
98
|
+
const parsed = Number(value);
|
|
99
|
+
if (!Number.isFinite(parsed) || parsed < 0)
|
|
100
|
+
throw invalidInput(`${name} must be a non-negative number`);
|
|
101
|
+
return parsed;
|
|
102
|
+
}
|
|
103
|
+
import { randomUUID } from "node:crypto";
|
|
104
|
+
import { stat } from "node:fs/promises";
|
|
105
|
+
import path from "node:path";
|
|
106
|
+
import { SCHEMA_VERSION, invalidInput } from "../foundation/index.js";
|
|
107
|
+
import { assertExactLocalGitEvalReference, parseHarnessReference } from "../revisions/index.js";
|
|
108
|
+
import { workspaceDigest } from "../workspaces/index.js";
|
|
109
|
+
//# sourceMappingURL=request.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"request.js","sourceRoot":"","sources":["../../../src/evals/request.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,MAAM,uBAAuB,GAAG,EAAE,GAAG,EAAE,GAAG,KAAK,CAAC;AACvD,MAAM,CAAC,MAAM,6BAA6B,GAAG,EAAE,GAAG,EAAE,GAAG,KAAK,CAAC;AAE7D,MAAM,sBAAsB,GAAqC;IAC/D,KAAK,EAAE,4CAA4C;IACnD,QAAQ,EAAE,gCAAgC;CAC3C,CAAC;AAEF,MAAM,UAAU,SAAS;IACvB,OAAO,QAAQ,UAAU,EAAE,CAAC,UAAU,CAAC,GAAG,EAAE,EAAE,CAAC,EAAY,CAAC;AAC9D,CAAC;AAgBD,MAAM,CAAC,KAAK,UAAU,mBAAmB,CAAC,KAAuB;IAC/D,IAAI,CAAC,KAAK,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC;QAAE,MAAM,YAAY,CAAC,oCAAoC,CAAC,CAAC;IAC1H,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC;QACtB,gBAAgB,EAAE,SAAS,EAAE,SAAS,EAAE,aAAa,EAAE,OAAO,EAAE,UAAU;QAC1E,gBAAgB,EAAE,YAAY,EAAE,kBAAkB,EAAE,YAAY,EAAE,UAAU;KAC7E,CAAC,CAAC;IACH,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC;IAC3E,IAAI,UAAU;QAAE,MAAM,YAAY,CAAC,+BAA+B,UAAU,EAAE,CAAC,CAAC;IAChF,IAAI,KAAK,CAAC,cAAc,KAAK,SAAS,IAAI,KAAK,CAAC,cAAc,KAAK,cAAc;QAAE,MAAM,YAAY,CAAC,+BAA+B,KAAK,CAAC,cAAc,EAAE,CAAC,CAAC;IAC7J,IAAI,CAAC,KAAK,CAAC,OAAO,IAAI,QAAQ,CAAC,KAAK,QAAQ;QAAE,MAAM,YAAY,CAAC,qDAAqD,CAAC,CAAC;IACxH,IAAI,OAAO,KAAK,CAAC,OAAO,KAAK,QAAQ,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,IAAI,EAAE;QAAE,MAAM,YAAY,CAAC,oCAAoC,CAAC,CAAC;IACzH,IAAI,OAAO,KAAK,CAAC,WAAW,KAAK,QAAQ,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,IAAI,EAAE;QAAE,MAAM,YAAY,CAAC,wCAAwC,CAAC,CAAC;IACrI,MAAM,SAAS,GAAG,qBAAqB,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC;IAC3D,IAAI,SAAS,CAAC,QAAQ,CAAC,IAAI,KAAK,WAAW,EAAE,CAAC;QAC5C,MAAM,YAAY,CAAC,yEAAyE,CAAC,CAAC;IAChG,CAAC;IACD,gCAAgC,CAAC,SAAS,CAAC,CAAC;IAC5C,MAAM,SAAS,GAAG,MAAM,yBAAyB,CAAC,KAAK,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC,CAAC;IACxE,MAAM,QAAQ,GAAG,eAAe,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,EAAE,UAAU,CAAC,CAAC;IAClE,MAAM,aAAa,GAAG,eAAe,CAAC,KAAK,CAAC,cAAc,IAAI,CAAC,EAAE,gBAAgB,CAAC,CAAC;IACnF,MAAM,OAAO,GAAG,iBAAiB,CAAC,KAAK,CAAC,UAAU,IAAI,uBAAuB,EAAE,YAAY,CAAC,CAAC;IAC7F,MAAM,YAAY,GAAG,iBAAiB,CAAC,KAAK,CAAC,gBAAgB,IAAI,6BAA6B,EAAE,kBAAkB,CAAC,CAAC;IACpH,IAAI,KAAK,CAAC,KAAK,KAAK,SAAS,IAAI,OAAO,KAAK,CAAC,KAAK,KAAK,QAAQ;QAAE,MAAM,YAAY,CAAC,wBAAwB,CAAC,CAAC;IAC/G,IAAI,KAAK,CAAC,UAAU,KAAK,SAAS,IAAI,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,UAAU,CAAC,IAAI,KAAK,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,EAAE,CAAC;QACxI,MAAM,YAAY,CAAC,wCAAwC,CAAC,CAAC;IAC/D,CAAC;IACD,IAAI,KAAK,CAAC,QAAQ,KAAK,SAAS,IAAI,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,QAAQ,CAAC,IAAI,KAAK,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,EAAE,CAAC;QAClI,MAAM,YAAY,CAAC,sCAAsC,CAAC,CAAC;IAC7D,CAAC;IACD,MAAM,SAAS,GAAG,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,UAAU,CAAa,CAAC,CAAC,CAAC,EAAE,CAAC;IAC3F,MAAM,SAAS,GAAG,sBAAsB,CAAC,SAAS,CAAC,UAAU,CAAC,CAAC;IAC/D,IAAI,SAAS,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,SAAS,CAAC;QAAE,SAAS,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;IAC9E,OAAO;QACL,cAAc,EAAE,cAAc;QAC9B,OAAO,EAAE,QAAQ;QACjB,OAAO,EAAE,MAAM,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE;QACrC,WAAW,EAAE,SAAS,CAAC,SAAS;QAChC,KAAK,EAAE,OAAO,KAAK,CAAC,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE;QACzD,QAAQ;QACR,cAAc,EAAE,aAAa;QAC7B,UAAU,EAAE,OAAO;QACnB,gBAAgB,EAAE,YAAY;QAC9B,UAAU,EAAE,SAAS;QACrB,QAAQ,EAAE,CAAC,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,QAAoB,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;QACvF,YAAY,EAAE,SAAS,CAAC,YAAY;QACpC,kBAAkB,EAAE,SAAS,CAAC,kBAAkB;KACjD,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,yBAAyB,CAAC,OAAe;IAC7D,MAAM,GAAG,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC;IAC3B,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC;IAChC,IAAI,CAAC;QACH,IAAI,CAAC,MAAM,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,WAAW,EAAE,EAAE,CAAC;YACtC,OAAO;gBACL,YAAY,EAAE,SAAS,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE;gBAC7C,kBAAkB,EAAE,MAAM,eAAe,CAAC,KAAK,EAAE,EAAE,gBAAgB,EAAE,IAAI,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,EAAE,CAAC;aAC1F,CAAC;QACJ,CAAC;IACH,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,IAAK,KAA+B,CAAC,IAAI,KAAK,QAAQ;YAAE,MAAM,KAAK,CAAC;IACtE,CAAC;IACD,MAAM,SAAS,GAAG,GAAG,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;IACvC,IAAI,SAAS,IAAI,CAAC,IAAI,SAAS,KAAK,GAAG,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACnD,MAAM,YAAY,CAAC,wEAAwE,CAAC,CAAC;IAC/F,CAAC;IACD,MAAM,WAAW,GAAG,GAAG,CAAC,KAAK,CAAC,CAAC,EAAE,SAAS,CAAC,CAAC;IAC5C,MAAM,QAAQ,GAAG,GAAG,CAAC,KAAK,CAAC,SAAS,GAAG,CAAC,CAAC,CAAC;IAC1C,IAAI,QAAQ,CAAC,WAAW,EAAE,KAAK,QAAQ;QAAE,MAAM,YAAY,CAAC,mCAAmC,CAAC,CAAC;IACjG,OAAO,EAAE,YAAY,EAAE,WAAW,EAAE,kBAAkB,EAAE,QAAQ,EAAE,CAAC;AACrE,CAAC;AAED,SAAS,eAAe,CAAC,KAAc,EAAE,IAAY;IACnD,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC;IAC7B,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,MAAM,CAAC,IAAI,MAAM,IAAI,CAAC;QAAE,MAAM,YAAY,CAAC,GAAG,IAAI,6BAA6B,CAAC,CAAC;IACvG,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,iBAAiB,CAAC,KAAc,EAAE,IAAY;IACrD,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC;IAC7B,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,MAAM,GAAG,CAAC;QAAE,MAAM,YAAY,CAAC,GAAG,IAAI,gCAAgC,CAAC,CAAC;IACxG,OAAO,MAAM,CAAC;AAChB,CAAC;AACD,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AACzC,OAAO,EAAE,IAAI,EAAE,MAAM,kBAAkB,CAAC;AACxC,OAAO,IAAI,MAAM,WAAW,CAAC;AAE7B,OAAO,EAAE,cAAc,EAAE,YAAY,EAAE,MAAM,wBAAwB,CAAC;AACtE,OAAO,EAAE,gCAAgC,EAAE,qBAAqB,EAAE,MAAM,uBAAuB,CAAC;AAChG,OAAO,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC"}
|
|
@@ -0,0 +1,227 @@
|
|
|
1
|
+
import path from "node:path";
|
|
2
|
+
import { resolveHarness } from "../artifacts/index.js";
|
|
3
|
+
import { HitchError, SCHEMA_VERSION, atomicWriteJSON, ensureDir, invalidInput, statePaths } from "../foundation/index.js";
|
|
4
|
+
import { parseHarnessReference } from "../revisions/index.js";
|
|
5
|
+
import { buildLocalGitTransport, lockedHarnessRef, runHarborBackend, verifyLocalGitTransport } from "../backends/index.js";
|
|
6
|
+
import { ensureControllerRuntime, writeRuntimeReference } from "../controller-runtime/index.js";
|
|
7
|
+
import { importEvalTrialRuns, validateEvalTrialReferences } from "./trial-import.js";
|
|
8
|
+
import { EvalEventSink } from "./events.js";
|
|
9
|
+
import { newEvalId, validateEvalRequest } from "./request.js";
|
|
10
|
+
export async function runEval({ evalId = newEvalId(), request, root, env = process.env, harborExecutable, signal, onEvent }) {
|
|
11
|
+
if (!root)
|
|
12
|
+
throw invalidInput("a Hitch state root is required for eval");
|
|
13
|
+
const normalized = await validateEvalRequest(request);
|
|
14
|
+
const evalDirectory = await ensureDir(path.join(statePaths(root).evals, evalId));
|
|
15
|
+
const startedAt = new Date();
|
|
16
|
+
const sink = new EvalEventSink(evalDirectory, evalId, onEvent);
|
|
17
|
+
await sink.open();
|
|
18
|
+
await atomicWriteJSON(path.join(evalDirectory, "request.json"), normalized);
|
|
19
|
+
let result;
|
|
20
|
+
let trialRefs = [];
|
|
21
|
+
try {
|
|
22
|
+
sink.emit({ type: "eval.started", backend: normalized.backend, dataset: normalized.dataset });
|
|
23
|
+
const requestedReference = parseHarnessReference(normalized.harness_ref);
|
|
24
|
+
const resolvedRevision = await resolveHarness(normalized.harness_ref, { root, env });
|
|
25
|
+
let localTransport;
|
|
26
|
+
if (resolvedRevision.source.type === "git" && resolvedRevision.source.registered !== true) {
|
|
27
|
+
const selector = requestedReference.selector;
|
|
28
|
+
if (selector.type !== "commit" || !selector.source?.explicit)
|
|
29
|
+
throw invalidInput("Harbor local Git eval requires an explicit git+file source");
|
|
30
|
+
if (resolvedRevision.revision.commit !== selector.value) {
|
|
31
|
+
throw new HitchError("local Git exact commit resolved to a different object", { code: "local_source_integrity_mismatch", exitCode: 12 });
|
|
32
|
+
}
|
|
33
|
+
localTransport = await buildLocalGitTransport({
|
|
34
|
+
evalDirectory,
|
|
35
|
+
resolvedRevision,
|
|
36
|
+
sourceDirectory: selector.source.local_path,
|
|
37
|
+
env,
|
|
38
|
+
...(signal ? { signal } : {}),
|
|
39
|
+
});
|
|
40
|
+
sink.emit({
|
|
41
|
+
type: "eval.local-source.prepared",
|
|
42
|
+
resolution_identity: localTransport.manifest.resolution_identity,
|
|
43
|
+
commit: localTransport.manifest.commit,
|
|
44
|
+
payload_sha256: localTransport.manifest.payload_sha256,
|
|
45
|
+
payload_bytes: localTransport.manifest.payload_bytes,
|
|
46
|
+
status: "verified",
|
|
47
|
+
});
|
|
48
|
+
}
|
|
49
|
+
// Phase 2: the shared, read-only, SHA-256-addressed controller runtime
|
|
50
|
+
// cache replaces the per-eval Hitch runtime copy (spec §4).
|
|
51
|
+
const controllerRuntime = await ensureControllerRuntime({ root });
|
|
52
|
+
const runtimeRefFile = await writeRuntimeReference(evalDirectory, controllerRuntime);
|
|
53
|
+
sink.emit({
|
|
54
|
+
type: "eval.controller-runtime",
|
|
55
|
+
runtime_id: controllerRuntime.runtime_id,
|
|
56
|
+
cache_hit: controllerRuntime.cache_hit,
|
|
57
|
+
reference: runtimeRefFile,
|
|
58
|
+
});
|
|
59
|
+
const plan = {
|
|
60
|
+
schema_version: SCHEMA_VERSION,
|
|
61
|
+
eval_id: evalId,
|
|
62
|
+
backend: "harbor",
|
|
63
|
+
candidate: {
|
|
64
|
+
id: "candidate-1",
|
|
65
|
+
requested_harness_ref: normalized.harness_ref,
|
|
66
|
+
harness_ref: lockedHarnessRef(resolvedRevision),
|
|
67
|
+
harness_id: resolvedRevision.harness_id,
|
|
68
|
+
revision_identity: resolvedRevision.identity,
|
|
69
|
+
model: normalized.model || null,
|
|
70
|
+
},
|
|
71
|
+
dataset: normalized.dataset,
|
|
72
|
+
benchmark_id: normalized.benchmark_id,
|
|
73
|
+
benchmark_revision: normalized.benchmark_revision,
|
|
74
|
+
attempts: normalized.attempts,
|
|
75
|
+
max_concurrent: normalized.max_concurrent,
|
|
76
|
+
controller_runtime: {
|
|
77
|
+
runtime_id: controllerRuntime.runtime_id,
|
|
78
|
+
manifest_digest: controllerRuntime.manifest_digest,
|
|
79
|
+
},
|
|
80
|
+
...(localTransport ? { local_source_transport: transportSummary(localTransport) } : {}),
|
|
81
|
+
created_at: new Date().toISOString(),
|
|
82
|
+
};
|
|
83
|
+
await atomicWriteJSON(path.join(evalDirectory, "resolution.json"), resolvedRevision);
|
|
84
|
+
await atomicWriteJSON(path.join(evalDirectory, "plan.json"), plan);
|
|
85
|
+
sink.emit({
|
|
86
|
+
type: "eval.planned",
|
|
87
|
+
harness: resolvedRevision.harness_id,
|
|
88
|
+
revision_identity: resolvedRevision.identity,
|
|
89
|
+
attempts: normalized.attempts,
|
|
90
|
+
});
|
|
91
|
+
if (localTransport) {
|
|
92
|
+
await verifyLocalGitTransport(localTransport, {
|
|
93
|
+
expected: {
|
|
94
|
+
harnessId: resolvedRevision.harness_id,
|
|
95
|
+
resolutionIdentity: resolvedRevision.identity,
|
|
96
|
+
commit: resolvedRevision.revision.commit,
|
|
97
|
+
},
|
|
98
|
+
env,
|
|
99
|
+
...(signal ? { signal } : {}),
|
|
100
|
+
});
|
|
101
|
+
}
|
|
102
|
+
const backendRun = await runHarborBackend({
|
|
103
|
+
evalId,
|
|
104
|
+
evalDirectory,
|
|
105
|
+
request: normalized,
|
|
106
|
+
root,
|
|
107
|
+
resolvedRevision,
|
|
108
|
+
runtimeDirectory: controllerRuntime.directory,
|
|
109
|
+
runtimeId: controllerRuntime.runtime_id,
|
|
110
|
+
...(localTransport ? { localTransport } : {}),
|
|
111
|
+
env,
|
|
112
|
+
...(harborExecutable !== undefined ? { harborExecutable } : {}),
|
|
113
|
+
...(signal ? { signal } : {}),
|
|
114
|
+
emit: (event) => sink.emit(event),
|
|
115
|
+
});
|
|
116
|
+
trialRefs = await importEvalTrialRuns({
|
|
117
|
+
root,
|
|
118
|
+
evalId,
|
|
119
|
+
evalDirectory,
|
|
120
|
+
request: normalized,
|
|
121
|
+
resolvedRevision,
|
|
122
|
+
benchmarkId: normalized.benchmark_id,
|
|
123
|
+
benchmarkRevision: normalized.benchmark_revision,
|
|
124
|
+
runtimeId: controllerRuntime.runtime_id,
|
|
125
|
+
rawResult: backendRun.rawResult,
|
|
126
|
+
});
|
|
127
|
+
await validateEvalTrialReferences(root, evalId, trialRefs, {
|
|
128
|
+
benchmarkId: normalized.benchmark_id,
|
|
129
|
+
benchmarkRevision: normalized.benchmark_revision,
|
|
130
|
+
});
|
|
131
|
+
const cancelled = signal?.aborted;
|
|
132
|
+
const localSourceFailure = localTransport ? localSourceBackendFailure(backendRun.rawResult) : false;
|
|
133
|
+
const succeeded = !cancelled && !localSourceFailure && backendRun.backend.process_exit_code === 0 && backendRun.rawResult !== null;
|
|
134
|
+
result = {
|
|
135
|
+
schema_version: SCHEMA_VERSION,
|
|
136
|
+
eval_id: evalId,
|
|
137
|
+
status: cancelled ? "cancelled" : succeeded ? "succeeded" : "failed",
|
|
138
|
+
exit_code: cancelled ? 9 : succeeded ? 0 : 13,
|
|
139
|
+
backend: backendRun.backend,
|
|
140
|
+
candidate: plan.candidate,
|
|
141
|
+
dataset: normalized.dataset,
|
|
142
|
+
benchmark_id: normalized.benchmark_id,
|
|
143
|
+
benchmark_revision: normalized.benchmark_revision,
|
|
144
|
+
trials: trialRefs,
|
|
145
|
+
summary: summarizeTrialRefs(trialRefs),
|
|
146
|
+
backend_summary: backendRun.summary,
|
|
147
|
+
...(localTransport ? { local_source_transport: transportSummary(localTransport) } : {}),
|
|
148
|
+
...(succeeded ? {} : {
|
|
149
|
+
error: {
|
|
150
|
+
code: cancelled ? "cancelled" : localSourceFailure ? "local_source_materialize_failed" : "harbor_failed",
|
|
151
|
+
message: cancelled
|
|
152
|
+
? "eval was cancelled"
|
|
153
|
+
: localSourceFailure
|
|
154
|
+
? "Harbor rejected the transported local Git source before candidate execution"
|
|
155
|
+
: backendRun.rawResult === null
|
|
156
|
+
? `Harbor exited without a result (code ${backendRun.backend.process_exit_code ?? "null"})`
|
|
157
|
+
: `Harbor exited with code ${backendRun.backend.process_exit_code ?? "null"}`,
|
|
158
|
+
},
|
|
159
|
+
}),
|
|
160
|
+
started_at: startedAt.toISOString(),
|
|
161
|
+
completed_at: new Date().toISOString(),
|
|
162
|
+
};
|
|
163
|
+
}
|
|
164
|
+
catch (error) {
|
|
165
|
+
const typed = error instanceof HitchError;
|
|
166
|
+
result = {
|
|
167
|
+
schema_version: SCHEMA_VERSION,
|
|
168
|
+
eval_id: evalId,
|
|
169
|
+
status: signal?.aborted ? "cancelled" : "failed",
|
|
170
|
+
exit_code: signal?.aborted ? 9 : typed ? error.exitCode : 12,
|
|
171
|
+
error: {
|
|
172
|
+
code: signal?.aborted ? "cancelled" : typed ? error.code : "internal_error",
|
|
173
|
+
message: error?.message || String(error),
|
|
174
|
+
},
|
|
175
|
+
benchmark_id: normalized.benchmark_id,
|
|
176
|
+
benchmark_revision: normalized.benchmark_revision,
|
|
177
|
+
trials: trialRefs,
|
|
178
|
+
started_at: startedAt.toISOString(),
|
|
179
|
+
completed_at: new Date().toISOString(),
|
|
180
|
+
};
|
|
181
|
+
}
|
|
182
|
+
await atomicWriteJSON(path.join(evalDirectory, "result.json"), result);
|
|
183
|
+
sink.emit({ type: result.status === "succeeded" ? "eval.completed" : "eval.failed", status: result.status, exit_code: result.exit_code, error: result.error });
|
|
184
|
+
await sink.close();
|
|
185
|
+
return result;
|
|
186
|
+
}
|
|
187
|
+
function summarizeTrialRefs(trials) {
|
|
188
|
+
const valid = trials.filter((trial) => trial.observation_status === "valid" && typeof trial.reward === "number");
|
|
189
|
+
const rewards = valid.map((trial) => trial.reward);
|
|
190
|
+
const aggregate = rewards.length
|
|
191
|
+
? {
|
|
192
|
+
count: rewards.length,
|
|
193
|
+
mean: rewards.reduce((sum, reward) => sum + reward, 0) / rewards.length,
|
|
194
|
+
min: Math.min(...rewards),
|
|
195
|
+
max: Math.max(...rewards),
|
|
196
|
+
}
|
|
197
|
+
: null;
|
|
198
|
+
return {
|
|
199
|
+
n_trials: trials.length,
|
|
200
|
+
n_completed: valid.length,
|
|
201
|
+
n_invalid: trials.length - valid.length,
|
|
202
|
+
primary_reward: aggregate?.mean ?? null,
|
|
203
|
+
rewards: aggregate ? { reward: aggregate } : {},
|
|
204
|
+
};
|
|
205
|
+
}
|
|
206
|
+
function localSourceBackendFailure(rawResult) {
|
|
207
|
+
if (!rawResult)
|
|
208
|
+
return false;
|
|
209
|
+
// The bridge uses this fixed, non-secret marker when setup cannot verify or
|
|
210
|
+
// materialize the uploaded source. Do not surface provider exception text in
|
|
211
|
+
// the durable Hitch error, because it may contain backend diagnostics.
|
|
212
|
+
return JSON.stringify(rawResult).includes("hitch-local-source-materialize:");
|
|
213
|
+
}
|
|
214
|
+
function transportSummary(transport) {
|
|
215
|
+
const manifest = transport.manifest;
|
|
216
|
+
return {
|
|
217
|
+
kind: manifest.kind,
|
|
218
|
+
resolution_identity: manifest.resolution_identity,
|
|
219
|
+
commit: manifest.commit,
|
|
220
|
+
tree: manifest.tree,
|
|
221
|
+
payload_sha256: manifest.payload_sha256,
|
|
222
|
+
payload_bytes: manifest.payload_bytes,
|
|
223
|
+
object_count: manifest.object_count,
|
|
224
|
+
file_count: manifest.file_count,
|
|
225
|
+
};
|
|
226
|
+
}
|
|
227
|
+
//# sourceMappingURL=service.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"service.js","sourceRoot":"","sources":["../../../src/evals/service.ts"],"names":[],"mappings":"AAAA,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAC;AACvD,OAAO,EAAE,UAAU,EAAE,cAAc,EAAE,eAAe,EAAE,SAAS,EAAE,YAAY,EAAY,UAAU,EAAE,MAAM,wBAAwB,CAAC;AACpI,OAAO,EAAE,qBAAqB,EAAE,MAAM,uBAAuB,CAAC;AAC9D,OAAO,EAAE,sBAAsB,EAAE,gBAAgB,EAAE,gBAAgB,EAAE,uBAAuB,EAAE,MAAM,sBAAsB,CAAC;AAE3H,OAAO,EAAE,uBAAuB,EAAE,qBAAqB,EAAE,MAAM,gCAAgC,CAAC;AAGhG,OAAO,EAAE,mBAAmB,EAAE,2BAA2B,EAAE,MAAM,mBAAmB,CAAC;AACrF,OAAO,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAC5C,OAAO,EAAE,SAAS,EAAE,mBAAmB,EAAE,MAAM,cAAc,CAAC;AAuB9D,MAAM,CAAC,KAAK,UAAU,OAAO,CAAC,EAAE,MAAM,GAAG,SAAS,EAAE,EAAE,OAAO,EAAE,IAAI,EAAE,GAAG,GAAG,OAAO,CAAC,GAAG,EAAE,gBAAgB,EAAE,MAAM,EAAE,OAAO,EAAkB;IACzI,IAAI,CAAC,IAAI;QAAE,MAAM,YAAY,CAAC,yCAAyC,CAAC,CAAC;IACzE,MAAM,UAAU,GAAG,MAAM,mBAAmB,CAAC,OAAO,CAAC,CAAC;IACtD,MAAM,aAAa,GAAG,MAAM,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC;IACjF,MAAM,SAAS,GAAG,IAAI,IAAI,EAAE,CAAC;IAC7B,MAAM,IAAI,GAAG,IAAI,aAAa,CAAC,aAAa,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC;IAC/D,MAAM,IAAI,CAAC,IAAI,EAAE,CAAC;IAClB,MAAM,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,cAAc,CAAC,EAAE,UAAU,CAAC,CAAC;IAC5E,IAAI,MAAkB,CAAC;IACvB,IAAI,SAAS,GAAkD,EAAE,CAAC;IAClE,IAAI,CAAC;QACH,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,cAAc,EAAE,OAAO,EAAE,UAAU,CAAC,OAAO,EAAE,OAAO,EAAE,UAAU,CAAC,OAAO,EAAE,CAAC,CAAC;QAC9F,MAAM,kBAAkB,GAAG,qBAAqB,CAAC,UAAU,CAAC,WAAW,CAAC,CAAC;QACzE,MAAM,gBAAgB,GAAG,MAAM,cAAc,CAAC,UAAU,CAAC,WAAW,EAAE,EAAE,IAAI,EAAE,GAAG,EAAE,CAAC,CAAC;QACrF,IAAI,cAAgD,CAAC;QACrD,IAAI,gBAAgB,CAAC,MAAM,CAAC,IAAI,KAAK,KAAK,IAAI,gBAAgB,CAAC,MAAM,CAAC,UAAU,KAAK,IAAI,EAAE,CAAC;YAC1F,MAAM,QAAQ,GAAG,kBAAkB,CAAC,QAAQ,CAAC;YAC7C,IAAI,QAAQ,CAAC,IAAI,KAAK,QAAQ,IAAI,CAAC,QAAQ,CAAC,MAAM,EAAE,QAAQ;gBAAE,MAAM,YAAY,CAAC,4DAA4D,CAAC,CAAC;YAC/I,IAAI,gBAAgB,CAAC,QAAQ,CAAC,MAAM,KAAK,QAAQ,CAAC,KAAK,EAAE,CAAC;gBACxD,MAAM,IAAI,UAAU,CAAC,uDAAuD,EAAE,EAAE,IAAI,EAAE,iCAAiC,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAC,CAAC;YAC3I,CAAC;YACD,cAAc,GAAG,MAAM,sBAAsB,CAAC;gBAC5C,aAAa;gBACb,gBAAgB;gBAChB,eAAe,EAAE,QAAQ,CAAC,MAAM,CAAC,UAAU;gBAC3C,GAAG;gBACH,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;aAC9B,CAAC,CAAC;YACH,IAAI,CAAC,IAAI,CAAC;gBACR,IAAI,EAAE,4BAA4B;gBAClC,mBAAmB,EAAE,cAAc,CAAC,QAAQ,CAAC,mBAAmB;gBAChE,MAAM,EAAE,cAAc,CAAC,QAAQ,CAAC,MAAM;gBACtC,cAAc,EAAE,cAAc,CAAC,QAAQ,CAAC,cAAc;gBACtD,aAAa,EAAE,cAAc,CAAC,QAAQ,CAAC,aAAa;gBACpD,MAAM,EAAE,UAAU;aACnB,CAAC,CAAC;QACL,CAAC;QACD,uEAAuE;QACvE,4DAA4D;QAC5D,MAAM,iBAAiB,GAA+B,MAAM,uBAAuB,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC;QAC9F,MAAM,cAAc,GAAG,MAAM,qBAAqB,CAAC,aAAa,EAAE,iBAAiB,CAAC,CAAC;QACrF,IAAI,CAAC,IAAI,CAAC;YACR,IAAI,EAAE,yBAAyB;YAC/B,UAAU,EAAE,iBAAiB,CAAC,UAAU;YACxC,SAAS,EAAE,iBAAiB,CAAC,SAAS;YACtC,SAAS,EAAE,cAAc;SAC1B,CAAC,CAAC;QACH,MAAM,IAAI,GAAG;YACX,cAAc,EAAE,cAAc;YAC9B,OAAO,EAAE,MAAM;YACf,OAAO,EAAE,QAAQ;YACjB,SAAS,EAAE;gBACT,EAAE,EAAE,aAAa;gBACjB,qBAAqB,EAAE,UAAU,CAAC,WAAW;gBAC7C,WAAW,EAAE,gBAAgB,CAAC,gBAAgB,CAAC;gBAC/C,UAAU,EAAE,gBAAgB,CAAC,UAAU;gBACvC,iBAAiB,EAAE,gBAAgB,CAAC,QAAQ;gBAC5C,KAAK,EAAE,UAAU,CAAC,KAAK,IAAI,IAAI;aAChC;YACD,OAAO,EAAE,UAAU,CAAC,OAAO;YAC3B,YAAY,EAAE,UAAU,CAAC,YAAY;YACrC,kBAAkB,EAAE,UAAU,CAAC,kBAAkB;YACjD,QAAQ,EAAE,UAAU,CAAC,QAAQ;YAC7B,cAAc,EAAE,UAAU,CAAC,cAAc;YACzC,kBAAkB,EAAE;gBAClB,UAAU,EAAE,iBAAiB,CAAC,UAAU;gBACxC,eAAe,EAAE,iBAAiB,CAAC,eAAe;aACnD;YACD,GAAG,CAAC,cAAc,CAAC,CAAC,CAAC,EAAE,sBAAsB,EAAE,gBAAgB,CAAC,cAAc,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YACvF,UAAU,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;SACrC,CAAC;QACF,MAAM,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,iBAAiB,CAAC,EAAE,gBAAgB,CAAC,CAAC;QACrF,MAAM,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,WAAW,CAAC,EAAE,IAAI,CAAC,CAAC;QACnE,IAAI,CAAC,IAAI,CAAC;YACR,IAAI,EAAE,cAAc;YACpB,OAAO,EAAE,gBAAgB,CAAC,UAAU;YACpC,iBAAiB,EAAE,gBAAgB,CAAC,QAAQ;YAC5C,QAAQ,EAAE,UAAU,CAAC,QAAQ;SAC9B,CAAC,CAAC;QACH,IAAI,cAAc,EAAE,CAAC;YACnB,MAAM,uBAAuB,CAAC,cAAc,EAAE;gBAC5C,QAAQ,EAAE;oBACR,SAAS,EAAE,gBAAgB,CAAC,UAAU;oBACtC,kBAAkB,EAAE,gBAAgB,CAAC,QAAQ;oBAC7C,MAAM,EAAE,gBAAgB,CAAC,QAAQ,CAAC,MAAgB;iBACnD;gBACD,GAAG;gBACH,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;aAC9B,CAAC,CAAC;QACL,CAAC;QACD,MAAM,UAAU,GAAG,MAAM,gBAAgB,CAAC;YACxC,MAAM;YACN,aAAa;YACb,OAAO,EAAE,UAAU;YACnB,IAAI;YACJ,gBAAgB;YAChB,gBAAgB,EAAE,iBAAiB,CAAC,SAAS;YAC7C,SAAS,EAAE,iBAAiB,CAAC,UAAU;YACvC,GAAG,CAAC,cAAc,CAAC,CAAC,CAAC,EAAE,cAAc,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC7C,GAAG;YACH,GAAG,CAAC,gBAAgB,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,gBAAgB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC/D,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC7B,IAAI,EAAE,CAAC,KAAK,EAAE,EAAE,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC;SAClC,CAAC,CAAC;QACH,SAAS,GAAG,MAAM,mBAAmB,CAAC;YACpC,IAAI;YACJ,MAAM;YACN,aAAa;YACb,OAAO,EAAE,UAAU;YACnB,gBAAgB;YAChB,WAAW,EAAE,UAAU,CAAC,YAAY;YACpC,iBAAiB,EAAE,UAAU,CAAC,kBAAkB;YAChD,SAAS,EAAE,iBAAiB,CAAC,UAAU;YACvC,SAAS,EAAE,UAAU,CAAC,SAAS;SAChC,CAAC,CAAC;QACH,MAAM,2BAA2B,CAAC,IAAI,EAAE,MAAM,EAAE,SAAS,EAAE;YACzD,WAAW,EAAE,UAAU,CAAC,YAAY;YACpC,iBAAiB,EAAE,UAAU,CAAC,kBAAkB;SACjD,CAAC,CAAC;QACH,MAAM,SAAS,GAAG,MAAM,EAAE,OAAO,CAAC;QAClC,MAAM,kBAAkB,GAAG,cAAc,CAAC,CAAC,CAAC,yBAAyB,CAAC,UAAU,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC;QACpG,MAAM,SAAS,GAAG,CAAC,SAAS,IAAI,CAAC,kBAAkB,IAAI,UAAU,CAAC,OAAO,CAAC,iBAAiB,KAAK,CAAC,IAAI,UAAU,CAAC,SAAS,KAAK,IAAI,CAAC;QACnI,MAAM,GAAG;YACP,cAAc,EAAE,cAAc;YAC9B,OAAO,EAAE,MAAM;YACf,MAAM,EAAE,SAAS,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,QAAQ;YACpE,SAAS,EAAE,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE;YAC7C,OAAO,EAAE,UAAU,CAAC,OAAO;YAC3B,SAAS,EAAE,IAAI,CAAC,SAAS;YACzB,OAAO,EAAE,UAAU,CAAC,OAAO;YAC3B,YAAY,EAAE,UAAU,CAAC,YAAY;YACrC,kBAAkB,EAAE,UAAU,CAAC,kBAAkB;YACjD,MAAM,EAAE,SAAS;YACjB,OAAO,EAAE,kBAAkB,CAAC,SAAS,CAAC;YACtC,eAAe,EAAE,UAAU,CAAC,OAAO;YACnC,GAAG,CAAC,cAAc,CAAC,CAAC,CAAC,EAAE,sBAAsB,EAAE,gBAAgB,CAAC,cAAc,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YACvF,GAAG,CAAC,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;gBACnB,KAAK,EAAE;oBACL,IAAI,EAAE,SAAS,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,kBAAkB,CAAC,CAAC,CAAC,iCAAiC,CAAC,CAAC,CAAC,eAAe;oBACxG,OAAO,EAAE,SAAS;wBAChB,CAAC,CAAC,oBAAoB;wBACtB,CAAC,CAAC,kBAAkB;4BAClB,CAAC,CAAC,6EAA6E;4BACjF,CAAC,CAAC,UAAU,CAAC,SAAS,KAAK,IAAI;gCAC7B,CAAC,CAAC,wCAAwC,UAAU,CAAC,OAAO,CAAC,iBAAiB,IAAI,MAAM,GAAG;gCAC3F,CAAC,CAAC,2BAA2B,UAAU,CAAC,OAAO,CAAC,iBAAiB,IAAI,MAAM,EAAE;iBAClF;aACF,CAAC;YACF,UAAU,EAAE,SAAS,CAAC,WAAW,EAAE;YACnC,YAAY,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;SACvC,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,MAAM,KAAK,GAAG,KAAK,YAAY,UAAU,CAAC;QAC1C,MAAM,GAAG;YACP,cAAc,EAAE,cAAc;YAC9B,OAAO,EAAE,MAAM;YACf,MAAM,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,QAAQ;YAChD,SAAS,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE;YAC5D,KAAK,EAAE;gBACL,IAAI,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,gBAAgB;gBAC3E,OAAO,EAAG,KAAe,EAAE,OAAO,IAAI,MAAM,CAAC,KAAK,CAAC;aACpD;YACD,YAAY,EAAE,UAAU,CAAC,YAAY;YACrC,kBAAkB,EAAE,UAAU,CAAC,kBAAkB;YACjD,MAAM,EAAE,SAAS;YACjB,UAAU,EAAE,SAAS,CAAC,WAAW,EAAE;YACnC,YAAY,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;SACvC,CAAC;IACJ,CAAC;IACD,MAAM,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,aAAa,CAAC,EAAE,MAAM,CAAC,CAAC;IACvE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,MAAM,CAAC,MAAM,KAAK,WAAW,CAAC,CAAC,CAAC,gBAAgB,CAAC,CAAC,CAAC,aAAa,EAAE,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,SAAS,EAAE,MAAM,CAAC,SAAS,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK,EAAE,CAAC,CAAC;IAC/J,MAAM,IAAI,CAAC,KAAK,EAAE,CAAC;IACnB,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,kBAAkB,CAAC,MAAqD;IAC/E,MAAM,KAAK,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,kBAAkB,KAAK,OAAO,IAAI,OAAO,KAAK,CAAC,MAAM,KAAK,QAAQ,CAAC,CAAC;IACjH,MAAM,OAAO,GAAG,KAAK,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,MAAgB,CAAC,CAAC;IAC7D,MAAM,SAAS,GAAG,OAAO,CAAC,MAAM;QAC9B,CAAC,CAAC;YACE,KAAK,EAAE,OAAO,CAAC,MAAM;YACrB,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,MAAM,EAAE,EAAE,CAAC,GAAG,GAAG,MAAM,EAAE,CAAC,CAAC,GAAG,OAAO,CAAC,MAAM;YACvE,GAAG,EAAE,IAAI,CAAC,GAAG,CAAC,GAAG,OAAO,CAAC;YACzB,GAAG,EAAE,IAAI,CAAC,GAAG,CAAC,GAAG,OAAO,CAAC;SAC1B;QACH,CAAC,CAAC,IAAI,CAAC;IACT,OAAO;QACL,QAAQ,EAAE,MAAM,CAAC,MAAM;QACvB,WAAW,EAAE,KAAK,CAAC,MAAM;QACzB,SAAS,EAAE,MAAM,CAAC,MAAM,GAAG,KAAK,CAAC,MAAM;QACvC,cAAc,EAAE,SAAS,EAAE,IAAI,IAAI,IAAI;QACvC,OAAO,EAAE,SAAS,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,SAAS,EAAE,CAAC,CAAC,CAAC,EAAE;KAChD,CAAC;AACJ,CAAC;AAED,SAAS,yBAAyB,CAAC,SAAyC;IAC1E,IAAI,CAAC,SAAS;QAAE,OAAO,KAAK,CAAC;IAC7B,4EAA4E;IAC5E,6EAA6E;IAC7E,uEAAuE;IACvE,OAAO,IAAI,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC,QAAQ,CAAC,iCAAiC,CAAC,CAAC;AAC/E,CAAC;AAED,SAAS,gBAAgB,CAAC,SAA+B;IACvD,MAAM,QAAQ,GAAG,SAAS,CAAC,QAAQ,CAAC;IACpC,OAAO;QACL,IAAI,EAAE,QAAQ,CAAC,IAAI;QACnB,mBAAmB,EAAE,QAAQ,CAAC,mBAAmB;QACjD,MAAM,EAAE,QAAQ,CAAC,MAAM;QACvB,IAAI,EAAE,QAAQ,CAAC,IAAI;QACnB,cAAc,EAAE,QAAQ,CAAC,cAAc;QACvC,aAAa,EAAE,QAAQ,CAAC,aAAa;QACrC,YAAY,EAAE,QAAQ,CAAC,YAAY;QACnC,UAAU,EAAE,QAAQ,CAAC,UAAU;KAChC,CAAC;AACJ,CAAC"}
|