Files
learnharness/tests/render-activity-dock.test.tsx
T
amosblomqvist de0c286311 feat: model + thinking control — the stats bar's picker, on pi-ai's catalog
The model-control wave (user-decided design, 2026-09-12): ONE controller
for the conversation agents (prober/teacher are functionally one agent);
background sub-agents stay config-pinned. pi-ai owns the registry — the
shipped 366-model OpenRouter catalog (cost, context, modality,
thinkingLevelMap), getSupportedThinkingLevels/clampThinkingLevel for the
per-model thinking truth, and streamSimple's automatic clamp. No second
registry to drift out of sync.

Seams:
- config.ts: ReasoningEffort retired for pi's ThinkingLevel
  (off|minimal|low|medium|high|xhigh|max). GLM roles go honest at "low" —
  pi's catalog says GLM-5.3-flash accepts exactly low|high|max, and the old
  "minimal" was silently clamped to low by streamSimple since the swap;
  nothing changes behaviorally. visualMaker → qwen/qwen3.8-flash (fixes
  the dead qwen/qwen3.8-max slug — the catalog only ships -0902; also the
  user's cost pick, 0.15/0.47 per-1M).
- src/agent/models.ts: piEngineFor(role, settings?) — provider+id resolved
  from the catalog, thinking level from settings ?? config, CLAMPED to the
  model's supported list (a settings change between models can never send
  an invalid level). PiEngine gains `level` (observability/tests).
- ChatRecord.settings {model?: {provider, id}, thinkingLevel?} — written
  by PUT /api/chats/[id]/settings (replace-whole, no updatedAt churn),
  read by chat-handler at turn composition: agents are stateless per turn,
  so the change is next-turn-effective by construction. The turns log
  still records the ACTUAL billed model (the honest witness).
- GET /api/models: the configured-provider catalog as client-safe DTOs
  (core/contracts/models.ts) + config defaults; a second provider joins
  the picker by registering it in piModels() — nothing else changes.
- PUT settings validation gate (src/server/model-catalog.ts): unknown
  provider/model or unsupported level → 400 carrying the supported list,
  never a silent clamp.
- scripts/sync-pricing.ts: regenerated PRICES/CONTEXT_LIMITS from pi-ai's
  shipped openrouter.json (offline; same source as the engine + picker).
  All 366 models priced; glm's stale 0.15/0.5 → pi's 0.075/0.25, limit
  1310720 → 1048576 — cost is inferred at read time, so history re-prices.
- UI (components/model-picker.tsx): stats-bar right side becomes model +
  thinking buttons; menus open ABOVE; fuzzy search over "provider id
  name" (core/fuzzy.ts — pi's /model pattern); model rows show name/id/
  price/context; the thinking menu lists exactly the chosen model's
  supported levels (GLM: low|high|max — no dead options, no "off").
  Model switch drops an incompatible pinned level (falls back to config,
  which clamps honestly) rather than carrying a 400.

Tests (350/350, +15): catalog projection + defaults, the settings gate,
pi thinking truth pinned against upgrades (glm low|high|max; minimal→low;
off→low — GLM cannot disable thinking), the engine factory's
override/fallback/clamp/throw matrix, fuzzy ranking, store round-trip,
and the re-priced folds (usage, stats bar, activity dock).

Live smoke (disposable driver, deleted): catalog route 366 + defaults;
glm+minimal → 400 + supported; turn 1 billed qwen/qwen3.8-flash under
the override; mid-session switch flipped turn 2 to glm; clear-overrides
restored the config pin. ALL CHECKS PASS on real OpenRouter.
2026-09-12 06:02:40 +03:00

174 lines
6.1 KiB
TypeScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import { renderToStaticMarkup } from "react-dom/server";
import { describe, it } from "node:test";
import assert from "node:assert/strict";
import { ActivityDock, formatElapsed, toBlobs, type ActivityFeed } from "../components/activity-dock.tsx";
import { formatTokens, formatUsd } from "../components/usage-format.ts";
// The dock's pure layer: folds, formatting, and blob order — then the SSR
// render (blobs, "—" doctrine, the abort affordance, the narrow badge).
const NOW = Date.parse("2026-09-09T12:00:00Z");
/** A full CallUsage (its fields are undefined-valued, not optional). */
function step(inputTokens: number, outputTokens: number) {
return { inputTokens, outputTokens, cacheReadTokens: undefined, cacheWriteTokens: undefined, reasoningTokens: undefined };
}
function job(over: Partial<ActivityFeed["jobs"][number]> = {}): ActivityFeed["jobs"][number] {
return {
id: "job-1",
kind: "research",
chatId: "c1",
toolCallId: "t1",
status: "running",
activity: 'web_search: "sigma.js v3"',
createdAt: new Date(NOW - 45_000).toISOString(),
updatedAt: new Date(NOW - 1000).toISOString(),
title: "Limits lesson",
...over,
};
}
describe("activity dock pure layer", () => {
it("merges jobs + extractions oldest-first (the stable blob order)", () => {
const feed: ActivityFeed = {
jobs: [job({ id: "new", createdAt: new Date(NOW - 10_000).toISOString() })],
extractions: [{ chatId: "cx", title: "Old lesson", phase: "auditing", startedAt: NOW - 100_000 }],
};
const blobs = toBlobs(feed);
assert.equal(blobs.length, 2);
const first = blobs[0]!;
assert.equal(first.id, "cx"); // extraction started first
assert.equal(first.process, "extraction");
assert.equal(first.phase, "auditing");
assert.equal(blobs[1]!.id, "new");
});
it("folds cost with jobCost; absent usage is null, never zero", () => {
const priced = job({
usage: {
model: "z-ai/glm-5.3-flash",
inputTokens: 10_000,
outputTokens: 1000,
cacheReadTokens: 5000,
cacheWriteTokens: 0,
reasoningTokens: 0,
steps: 2,
},
});
const blob = toBlobs({ jobs: [priced], extractions: [] })[0]!;
// (10000-5000)/1e6*0.075 + 5000/1e6*0.015 + 1000/1e6*0.25 = 0.0007
// (pi-ai catalog prices)
assert.ok(blob.cost !== null && blob.cost > 0.0006 && blob.cost < 0.0008);
const absent = toBlobs({ jobs: [job()], extractions: [] })[0]!;
assert.equal(absent.cost, null);
});
it("context line = LAST step's tokens vs the model limit, never a cumulative sum", () => {
const base = {
model: "z-ai/glm-5.3-flash",
inputTokens: 400_000, // running total (much bigger — must NOT feed the line)
outputTokens: 8_000,
cacheReadTokens: 0,
cacheWriteTokens: 0,
reasoningTokens: 0,
steps: 5,
};
const withLast = job({
usage: { ...base, lastStep: step(90_000, 2_000) },
});
const blob = toBlobs({ jobs: [withLast], extractions: [] })[0]!;
assert.equal(blob.context?.used, 92_000);
assert.equal(blob.context?.limit, 1048576);
// No lastStep yet (capture pending) → "—", even though usage exists.
const stripped = toBlobs({ jobs: [job({ usage: { ...base } })], extractions: [] })[0]!;
assert.equal(stripped.context, null);
// Unknown model → no limit → "—".
const unknownModel = toBlobs({
jobs: [job({ usage: { ...base, model: "nope/unknown", lastStep: step(5, 5) } })],
extractions: [],
})[0]!;
assert.equal(unknownModel.context, null);
});
it("extraction blobs carry no money/context (—, the doctrine)", () => {
const blob = toBlobs({
jobs: [],
extractions: [{ chatId: "cx", title: "T", phase: "extracting", startedAt: NOW }],
})[0]!;
assert.equal(blob.cost, null);
assert.equal(blob.context, null);
});
it("formatElapsed ticks compactly", () => {
assert.equal(formatElapsed(45_000), "45s");
assert.equal(formatElapsed(184_000), "3m 04s");
assert.equal(formatElapsed(4_320_000), "1h 12m");
});
it("formatUsd keeps four decimals under $1, two above", () => {
assert.equal(formatUsd(0.0014), "$0.0014");
assert.equal(formatUsd(1.5), "$1.50");
});
it("formatTokens trims (never 38.0k)", () => {
assert.equal(formatTokens(38_000), "38k");
assert.equal(formatTokens(1_234_567), "1.2M");
assert.equal(formatTokens(950), "950");
});
});
describe("activity dock render", () => {
it("renders NOTHING when idle (no jobs, no extractions)", () => {
assert.equal(renderToStaticMarkup(<ActivityDock feed={{ jobs: [], extractions: [] }} />), "");
});
it("renders one blob per running process with cost, elapsed, and the abort affordance", () => {
const html = renderToStaticMarkup(
<ActivityDock
feed={{
jobs: [job({ id: "j1", kind: "demo", activity: "building sandbox…", title: "Chain-rule demo" })],
extractions: [{ chatId: "cx", title: "Old lesson", phase: "merging", startedAt: NOW - 60_000 }],
}}
/>,
);
assert.match(html, /data-testid="activity-dock"/);
assert.match(html, /data-testid="activity-blob-job"/);
assert.match(html, /data-testid="activity-blob-extraction"/);
assert.match(html, /Chain-rule demo/);
assert.match(html, /building sandbox…/);
assert.match(html, /Extraction — merging/);
// Unpriced job → "—", never $0.0000 (doctrine); both blobs show it.
assert.match(html, /activity-blob-cost[^>]*>—</);
assert.match(html, /data-testid="activity-abort"/, "END on the blob face, popover-independent");
assert.match(html, /\bEnd\b/, "the explicit END button");
assert.match(html, /Background activity/);
// Total cost: nothing priced → "—" in the header.
assert.match(html, /activity-total-cost[^>]*>—</);
});
it("shows total running cost when jobs are priced", () => {
const usage = {
model: "z-ai/glm-5.3-flash",
inputTokens: 10_000,
outputTokens: 1000,
cacheReadTokens: 0,
cacheWriteTokens: 0,
reasoningTokens: 0,
steps: 1,
};
const html = renderToStaticMarkup(
<ActivityDock
feed={{
jobs: [
job({ id: "a", usage }),
job({ id: "b", usage, createdAt: new Date(NOW - 20_000).toISOString() }),
],
extractions: [],
}}
/>,
);
// two identical priced jobs → 2 × $0.001 = $0.0020 total (pi prices)
assert.match(html, /activity-total-cost[^>]*>\$0\.0020</);
});
});