Files
PaperClipAI/tests/runner-e2e/billing.ts
T
DottaandPaperclip b648d8cdda fix(evals): support explicit Grok qualification workflows (#13878)
## Thinking Path

> - Paperclip manages AI agents and their provider connections.
> - Product E2E checks real tasks through the browser, server, and
runner.
> - Grok qualification needs separate API-key and subscription evidence.
> - Product subscription tests and direct Grok protocol evals need
explicit credential delivery.
> - This change supplies each credential only to its selected profile
and prepares the pinned binary.
> - Maintainer authorization and protected-environment gates remain
required.

## Linked Issues or Issue Description

Refs #13845, #13847, #13850, #13882.

The Grok feature branch has a manual subscription qualification profile.
The trusted master workflow must admit its selected credential and
prepare the same verified binary and artifact verifier as the API
profile. Direct protocol evals also need the selected xAI key and pinned
Grok binary. These prerequisites do not register or schedule the new
profiles on master.

## What Changed

- Deliver `GROK_AUTH_JSON` from the protected paid environment only when
the selected profile requests that credential.
- Install the checksum-verified Grok binary for the local subscription
profile.
- Prepare the pinned artifact verifier for the manual subscription
suite.
- Extend workflow security assertions to cover the new credential and
profile.
- Add the ACPX Grok credential mapping to the trusted-master catalog,
then deliver only the selected `XAI_API_KEY` to direct protocol cells
and install the target’s checksum-verified Grok binary before packaging.
- Allow a direct-protocol concurrency override from two cases up to the
existing configured ceiling; it can only lower concurrency.
- Document the Grok protocol workflow and its API-only credential
boundary.
- Render missing LLM usage and cost as Unavailable, and label partial
observations with coverage. Preserve raw records, grades, and actual
zero costs.
- Preserve measured campaign source metadata during report regeneration
instead of inheriting the renderer checkout or CI event; skip empty
legacy source records when recovering older provenance.

## Verification

- Latest commit `05d05801477104c8155977bbbe3e119a5241f960`: all 54
reported checks successful, two intentional skips, Greptile 5/5, and
zero unresolved review threads. [CI
run](https://github.com/paperclipai/paperclip/actions/runs/35890978288).

- After merging current master, all 17 workflow security/image tests and
23 catalog/workflow policy tests passed. The trusted catalog also
generated all 39 pinned Grok cells with `XAI_API_KEY` and one case per
shard. The new policy tests execute the concurrency guard against valid,
out-of-range, and malformed values.
- The Grok branch separately passed 450 Product harness unit tests,
including private company credential staging, cleanup, and
token-fragment redaction.
- The fresh-login native subscription smoke passed three repetitions of
tool execution, session resume, restrictive permissions, and cleanup.
These are setup evidence; full subscription Product qualification
remains pending.
- All 72 focused report/billing/history/catalog tests and the Product
harness typecheck passed for the report-display change. The initial
sandbox run could not open the tsx IPC socket; the permitted rerun
passed. A zero-provider-call replay of the actual 16-cell Grok campaign
preserved all result records, grades, timing, and source provenance
while correcting missing usage labels.
- Review the thirteen-file diff. Provider credentials still enter only
the selected paid-test step; default-branch, numeric-actor, and
environment restrictions are unchanged.

## Risks

This admits a refreshable subscription credential to explicitly selected
trusted tests. Store it only in `runner-e2e-paid`, use a test login, and
remove it after qualification. Unselected profiles receive an empty
value. Pull requests cannot trigger the paid workflow. This PR changes
no fleet admission or actor allowlist.

## Model Used

OpenAI GPT-6 through Codex, with tool use and code execution. The exact
serving model identifier and context-window size are not exposed in this
session.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [x] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

---------

Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-09-23 17:05:52 -05:00

416 lines
13 KiB
TypeScript

import type {
RunnerE2EBillingSummary,
RunnerE2EResult,
RunnerE2ERuntimeUsage,
RunnerEnvironmentId,
} from "./types.js";
// Public list prices are deliberately versioned here instead of being treated
// as provider-reported charges. Credits, discounts, and Daytona's first 5 GiB
// storage allowance can make the invoice amount lower than this estimate.
export const DAYTONA_LIST_PRICING = {
asOf: "2026-08-27",
url: "https://www.daytona.io/pricing",
cpuCoreHourUsd: 0.0504,
memoryGiBHourUsd: 0.0162,
diskGiBHourUsd: 0.000108,
} as const;
export interface RuntimeLeaseUsageInput {
acquiredAt?: string | Date | null;
releasedAt?: string | Date | null;
updatedAt?: string | Date | null;
metadata?: Record<string, unknown> | null;
}
export interface AgentRunUsageInput {
startedAt?: string | Date | null;
finishedAt?: string | Date | null;
}
export interface CampaignBillingSummary {
judge?: import("./types.js").RunnerE2EJudgeBillingSummary;
testCount: number;
agentRunDurationMs: number;
leaseDurationMs: number;
llm: RunnerE2EBillingSummary["llm"];
reportedLlmCostUsd: number;
estimatedRuntimeCostUsd: number;
observedAndEstimatedCostUsd: number | null;
testsWithCompleteBilling: number;
}
/** Render observed subtotals without presenting absent measurements as zero. */
export function billingCoverageLabel(value: string, covered: number, total: number): string {
if (!Number.isInteger(covered) || !Number.isInteger(total) || covered <= 0 || total <= 0 || covered > total) {
return "Unavailable";
}
return covered < total ? `${value} (partial: ${covered}/${total} runs)` : value;
}
function record(value: unknown): Record<string, unknown> {
return value && typeof value === "object" && !Array.isArray(value)
? (value as Record<string, unknown>)
: {};
}
function finiteNumber(value: unknown): number | undefined {
return typeof value === "number" && Number.isFinite(value) && value >= 0
? value
: undefined;
}
function firstNumber(source: Record<string, unknown>, keys: readonly string[]) {
for (const key of keys) {
const value = finiteNumber(source[key]);
if (value !== undefined) return value;
}
return undefined;
}
function usageMeasurement(usage: Record<string, unknown>) {
const candidates = [
record(usage.runDelta),
record(usage.total),
record(usage.cumulative),
usage,
];
return (
candidates.find(
(candidate) =>
firstNumber(candidate, [
"inputTokens",
"input",
"promptTokens",
"outputTokens",
"output",
"completionTokens",
]) !== undefined,
) ?? usage
);
}
function usageCostUsd(usage: Record<string, unknown>) {
const measurement = usageMeasurement(usage);
const direct =
firstNumber(usage, [
"cacheAdjustedCostUsd",
"costUsd",
"providerCostUsd",
]) ??
firstNumber(measurement, [
"cacheAdjustedCostUsd",
"costUsd",
"providerCostUsd",
]);
if (direct !== undefined) return direct;
const cost = record(usage.cost);
const currency =
typeof cost.currency === "string" ? cost.currency.toUpperCase() : "USD";
if (currency !== "USD") return undefined;
return firstNumber(cost, ["amount", "total"]);
}
function usageEntries(
rawUsage: Record<string, unknown> | null | undefined,
runCount: number,
) {
const runs = Array.isArray(rawUsage?.runs) ? rawUsage.runs : null;
if (runs) {
const entries = runs.map((entry) => {
const candidate = record(entry);
const usage = candidate.usage;
return usage && typeof usage === "object" && !Array.isArray(usage)
? (usage as Record<string, unknown>)
: null;
});
return [
...entries,
...Array.from(
{ length: Math.max(0, runCount - entries.length) },
() => null,
),
];
}
return Array.from({ length: Math.max(1, runCount) }, (_, index) =>
index === 0 && rawUsage ? rawUsage : null,
);
}
function durationBetween(
startedAt: string | Date | null | undefined,
finishedAt: string | Date | null | undefined,
) {
const started = startedAt ? new Date(startedAt).getTime() : Number.NaN;
const finished = finishedAt ? new Date(finishedAt).getTime() : Number.NaN;
return Number.isFinite(started) && Number.isFinite(finished)
? Math.max(0, finished - started)
: 0;
}
function resourceValue(
metadata: Record<string, unknown> | null | undefined,
key: "cpu" | "memory" | "disk",
) {
return finiteNumber(metadata?.[key]);
}
export function buildRuntimeUsage(input: {
environmentId: RunnerEnvironmentId;
runs: readonly AgentRunUsageInput[];
leases?: readonly RuntimeLeaseUsageInput[];
fallbackFinishedAt?: string | Date;
}): RunnerE2ERuntimeUsage {
const agentRunDurationMs = input.runs.reduce(
(total, run) => total + durationBetween(run.startedAt, run.finishedAt),
0,
);
if (input.environmentId === "local") {
return {
provider: "local",
agentRunDurationMs,
leaseDurationMs: null,
leaseCount: 0,
costStatus: "not_metered",
costSource: "local_not_metered",
};
}
const leases = input.leases ?? [];
let leaseDurationMs = 0;
let estimatedListCostUsd = 0;
let resourcesComplete = leases.length > 0;
const resourceRows: Array<{ cpu: number; memory: number; disk: number }> = [];
for (const lease of leases) {
const finishedAt =
lease.releasedAt ?? input.fallbackFinishedAt ?? lease.updatedAt ?? null;
const durationMs = durationBetween(lease.acquiredAt, finishedAt);
leaseDurationMs += durationMs;
const cpu = resourceValue(lease.metadata, "cpu");
const memory = resourceValue(lease.metadata, "memory");
const disk = resourceValue(lease.metadata, "disk");
if (cpu === undefined || memory === undefined || disk === undefined) {
resourcesComplete = false;
continue;
}
resourceRows.push({ cpu, memory, disk });
const hours = durationMs / 3_600_000;
estimatedListCostUsd +=
hours *
(cpu * DAYTONA_LIST_PRICING.cpuCoreHourUsd +
memory * DAYTONA_LIST_PRICING.memoryGiBHourUsd +
disk * DAYTONA_LIST_PRICING.diskGiBHourUsd);
}
const commonResource = (key: "cpu" | "memory" | "disk") => {
const values = new Set(resourceRows.map((row) => row[key]));
return values.size === 1 ? resourceRows[0]?.[key] : undefined;
};
return {
provider: "daytona",
agentRunDurationMs,
leaseDurationMs: leases.length > 0 ? leaseDurationMs : null,
leaseCount: leases.length,
...(commonResource("cpu") === undefined
? {}
: { cpuCores: commonResource("cpu") }),
...(commonResource("memory") === undefined
? {}
: { memoryGiB: commonResource("memory") }),
...(commonResource("disk") === undefined
? {}
: { diskGiB: commonResource("disk") }),
...(resourcesComplete ? { estimatedListCostUsd } : {}),
costStatus: resourcesComplete ? "estimated" : "unavailable",
costSource: resourcesComplete
? "daytona_public_list_price"
: "provider_cost_unavailable",
...(resourcesComplete
? {
pricingAsOf: DAYTONA_LIST_PRICING.asOf,
pricingUrl: DAYTONA_LIST_PRICING.url,
}
: {}),
};
}
export function fallbackRuntimeUsage(
result: Pick<
RunnerE2EResult,
"environmentId" | "durationMs" | "runtimeUsage"
>,
) {
if (result.runtimeUsage) return result.runtimeUsage;
return {
provider: result.environmentId,
agentRunDurationMs: result.durationMs,
leaseDurationMs: null,
leaseCount: 0,
costStatus:
result.environmentId === "local" ? "not_metered" : "unavailable",
costSource:
result.environmentId === "local"
? "local_not_metered"
: "provider_cost_unavailable",
} satisfies RunnerE2ERuntimeUsage;
}
export function summarizeExecutionBilling(
result: Pick<
RunnerE2EResult,
"runIds" | "usage" | "environmentId" | "durationMs" | "runtimeUsage" | "firstTaskQuality"
>,
): RunnerE2EBillingSummary {
const requestedRunCount = Math.max(result.runIds?.length ?? 0, 1);
const entries = usageEntries(result.usage, requestedRunCount);
let inputTokens = 0;
let outputTokens = 0;
let cachedInputTokens = 0;
let runsWithTokenUsage = 0;
let runsWithReportedCost = 0;
let reportedCostUsd = 0;
for (const usage of entries) {
if (!usage) continue;
const measurement = usageMeasurement(usage);
const input =
firstNumber(measurement, ["inputTokens", "input", "promptTokens"]) ?? 0;
const output =
firstNumber(measurement, [
"outputTokens",
"output",
"completionTokens",
]) ?? 0;
const cached =
firstNumber(measurement, [
"cachedInputTokens",
"cacheReadTokens",
"cachedReadTokens",
]) ?? 0;
inputTokens += input;
outputTokens += output;
cachedInputTokens += cached;
if (input > 0 || output > 0 || cached > 0) runsWithTokenUsage += 1;
const costUsd = usageCostUsd(usage);
if (costUsd !== undefined && (costUsd > 0 || input > 0 || output > 0)) {
runsWithReportedCost += 1;
reportedCostUsd += costUsd;
}
}
const runCount = Math.max(requestedRunCount, entries.length);
const costStatus =
runsWithReportedCost === runCount
? "reported"
: runsWithReportedCost > 0
? "partial"
: runsWithTokenUsage > 0
? "unpriced"
: "unavailable";
const runtime = fallbackRuntimeUsage(result);
const estimatedRuntimeCostUsd = runtime.estimatedListCostUsd ?? 0;
const quality = result.firstTaskQuality;
const complete =
(!quality || quality.estimatedCostUsd !== null) &&
runsWithTokenUsage === runCount &&
runsWithReportedCost === runCount &&
runtime.costStatus !== "unavailable";
return {
llm: {
runCount,
runsWithTokenUsage,
runsWithReportedCost,
inputTokens,
outputTokens,
cachedInputTokens,
totalTokens: inputTokens + cachedInputTokens + outputTokens,
reportedCostUsd,
costStatus,
},
runtime,
reportedCostUsd,
estimatedRuntimeCostUsd,
...(quality ? { judge: { inputTokens: quality.inputTokens, outputTokens: quality.outputTokens, estimatedCostUsd: quality.estimatedCostUsd, reservedCostUsd: quality.reservedCostUsd } } : {}),
observedAndEstimatedCostUsd: quality?.estimatedCostUsd === null ? null : reportedCostUsd + estimatedRuntimeCostUsd + (quality?.estimatedCostUsd ?? 0),
complete,
};
}
export function aggregateCampaignBilling(
results: readonly RunnerE2EResult[],
): CampaignBillingSummary {
const summaries = results.map(summarizeExecutionBilling);
const runCount = summaries.reduce(
(total, summary) => total + summary.llm.runCount,
0,
);
const runsWithTokenUsage = summaries.reduce(
(total, summary) => total + summary.llm.runsWithTokenUsage,
0,
);
const runsWithReportedCost = summaries.reduce(
(total, summary) => total + summary.llm.runsWithReportedCost,
0,
);
const reportedLlmCostUsd = summaries.reduce(
(total, summary) => total + summary.reportedCostUsd,
0,
);
const estimatedRuntimeCostUsd = summaries.reduce(
(total, summary) => total + summary.estimatedRuntimeCostUsd,
0,
);
const judges = summaries.flatMap(summary => summary.judge ? [summary.judge] : []);
return {
...(judges.length ? { judge: {
attempts: judges.length,
inputTokens: judges.reduce((n, q) => n + (q.inputTokens ?? 0), 0),
outputTokens: judges.reduce((n, q) => n + (q.outputTokens ?? 0), 0),
estimatedCostUsd: judges.some(q => q.estimatedCostUsd === null) ? null : judges.reduce((n, q) => n + (q.estimatedCostUsd ?? 0), 0),
reservedCostUsd: judges.reduce((n, q) => n + q.reservedCostUsd, 0),
attemptsWithUnknownUsage: judges.filter(q => q.estimatedCostUsd === null).length,
} } : {}),
testCount: results.length,
agentRunDurationMs: summaries.reduce(
(total, summary) => total + summary.runtime.agentRunDurationMs,
0,
),
leaseDurationMs: summaries.reduce(
(total, summary) => total + (summary.runtime.leaseDurationMs ?? 0),
0,
),
llm: {
runCount,
runsWithTokenUsage,
runsWithReportedCost,
inputTokens: summaries.reduce(
(total, summary) => total + summary.llm.inputTokens,
0,
),
outputTokens: summaries.reduce(
(total, summary) => total + summary.llm.outputTokens,
0,
),
cachedInputTokens: summaries.reduce(
(total, summary) => total + summary.llm.cachedInputTokens,
0,
),
totalTokens: summaries.reduce(
(total, summary) => total + summary.llm.totalTokens,
0,
),
reportedCostUsd: reportedLlmCostUsd,
costStatus:
runsWithReportedCost === runCount
? "reported"
: runsWithReportedCost > 0
? "partial"
: runsWithTokenUsage > 0
? "unpriced"
: "unavailable",
},
reportedLlmCostUsd,
estimatedRuntimeCostUsd,
observedAndEstimatedCostUsd: summaries.some(s => s.observedAndEstimatedCostUsd === null) ? null : summaries.reduce((total, summary) => total + (summary.observedAndEstimatedCostUsd ?? 0), 0),
testsWithCompleteBilling: summaries.filter((summary) => summary.complete)
.length,
};
}