Files
learnharness/tests/phases.test.ts
T
amosblomqvist 740293c361 feat: custom message types for system turns — wake and verdict messages
The custom-message scheme (user-decided 2026-09-12): system-authored
user-role messages are NOT human messages — they carry marker data parts
only, the adapter owns their model-facing wording, and the UI renders
them under control. Nothing fake is ever persisted as human speech.

The inventory (all migrated):
- probe→teacher handoff wake: was a text-bearing user message rendering
  as a fake user bubble; now data-probe-handoff marker-only — the
  adapter flattens it to HANDOFF_TEXT (core/contracts/chat.ts), the UI
  renders a quiet 'Preparing your plan…' system line.
- plan accept/decline: was boilerplate text ('The plan is accepted…',
  'I'd like the plan revised…') masquerading as learner speech — the
  decline wasn't even marked; now data-verdict {accepted, at} on its own
  message, adapter flattens to verdictText(), UI renders the quiet
  verdict marker ('✓ Plan accepted' / '↻ Plan revision requested').
  sendVerdict sends TWO REAL MESSAGES when the learner typed a note
  (the note as a genuine user message, then the verdict) in ONE turn
  via setMessages + sendMessage() — no 409 dance, no double teacher turn.
- sub-agent steers: the abort steer gains its data-subagent-aborted
  marker part; the UI keys steer rendering off PARTS (artifact/aborted),
  not metadata — metadata stays as pump mechanics only.
- record-adapter: systemTextOf() flattens marker-only messages; steers
  keep their text (payload — briefs are content, not boilerplate).

phases.ts: plan→teach advances on data-verdict{accepted:true} only;
learner text alone can never sanction teaching (pinned). data-plan-
accepted retired, no back-compat (user-decided: wiped-data doctrine).

Pinned: record-adapter.test.ts (verdict/handoff flattening, note+verdict
order, steer payload kept, UI-only parts skipped), phases.test.ts
(verdict marker semantics, out-of-order, text-never-fakes), jobs.test.ts
(abort marker part), demos.test.tsx (badge renders off the part, not
metadata).

Live GLM smoke (disposable driver, deleted): probe→conclude→wake→plan→
decline-with-note (revised, still plan)→accept-with-note (teach) — ALL
CHECKS PASS, including marker-only handoff/verdict on the wire and zero
fake human speech in the transcript.

Gates: tsc, lint, build, 335/335.
2026-09-12 05:14:40 +03:00

140 lines
5.5 KiB
TypeScript

/**
* phases — the event-driven phase chain (docs/probe-teach-split.md §4).
*
* probe ends at the first CONCLUDED probe_done (output-available); planning
* ends at the LEARNER'S plan acceptance (a data-verdict part with accepted:true
* accept button's user message — plan_init only PRESENTS the map). No
* stored-tree seam: the phase is a pure fold over the transcript, so it can
* never stick (the old currentNodeId-based phaseOfTree read a field nothing
* ever set).
*/
import assert from "node:assert";
import test from "node:test";
import type { UIMessage } from "ai";
import { phaseAt3, phaseTimeline, type Phase3 } from "../src/core/phases.ts";
/** Text-only user message. */
function user(text: string): UIMessage {
return { id: `u_${text}`, role: "user", parts: [{ type: "text", text }] };
}
/** An assistant message carrying a tool part in the given state. */
function assistantTool(
name: string,
state: "input-available" | "output-available" | "output-error",
output?: unknown,
): UIMessage {
return {
id: `a_${name}_${state}_${Math.random().toString(36).slice(2, 6)}`,
role: "assistant",
parts: [{ type: `tool-${name}`, state, ...(output !== undefined ? { output } : {}) } as never],
};
}
const DONE_ASSESSMENT = { assessment: { strands: [{ strand: "s", assessment: "a" }] } };
const PLAN_OK = { ok: true, tree: { nodes: [{ id: "n1", title: "N" }], edges: [] } };
/** The verdict button's user message (the custom-message scheme): its own
* message carrying ONLY the data-verdict marker part — no human-authored
* text; the model-facing wording is the adapter's (record-adapter.ts). */
function verdict(accepted: boolean): UIMessage {
return {
id: `u_verdict_${Math.random().toString(36).slice(2, 6)}`,
role: "user",
parts: [
{ type: "data-verdict", id: "vd_1", data: { accepted, at: "2026-09-12T00:00:00Z" } } as never,
],
};
}
/** The accept button's message. */
function planAccepted(): UIMessage {
return verdict(true);
}
test("phaseAt3: probe until a concluded probe_done; plan until the learner ACCEPTS; teach after", () => {
const probe = phaseAt3([user("teach me calculus"), assistantTool("quiz", "output-available")]);
assert.equal(probe, "probe", "quiz answers alone never end the probe");
const concluded = phaseAt3([
user("teach me"),
assistantTool("probe_done", "output-available", DONE_ASSESSMENT),
]);
assert.equal(concluded, "plan", "a concluded probe_done ends the probe (planning begins)");
const presented = phaseAt3([
user("teach me"),
assistantTool("probe_done", "output-available", DONE_ASSESSMENT),
assistantTool("plan_init", "output-available", PLAN_OK),
]);
assert.equal(presented, "plan", "a successful plan_init PRESENTS the map — planning continues until the verdict");
const accepted = phaseAt3([
user("teach me"),
assistantTool("probe_done", "output-available", DONE_ASSESSMENT),
assistantTool("plan_init", "output-available", PLAN_OK),
planAccepted(),
]);
assert.equal(accepted, "teach", "the learner's acceptance ends planning");
const declined = phaseAt3([
user("teach me"),
assistantTool("probe_done", "output-available", DONE_ASSESSMENT),
assistantTool("plan_init", "output-available", PLAN_OK),
verdict(false),
assistantTool("plan_init", "output-available", PLAN_OK),
]);
assert.equal(declined, "plan", "a DECLINED verdict is a real event but stays in planning until a later acceptance");
const failedPlan = phaseAt3([
user("teach me"),
assistantTool("probe_done", "output-available", DONE_ASSESSMENT),
assistantTool("plan_init", "output-error", { ok: false, error: "bad mermaid" }),
planAccepted(),
]);
assert.equal(failedPlan, "teach", "acceptance sanctions teaching even after a failed init attempt — the verdict is his");
});
test("phaseAt3: the verdict marker is explicit — anonymous learner text never sanctions teaching", () => {
const plainText = phaseAt3([
user("teach me"),
assistantTool("probe_done", "output-available", DONE_ASSESSMENT),
assistantTool("plan_init", "output-available", PLAN_OK),
user("The plan is accepted — begin teaching."), // words alone, no marker
]);
assert.equal(plainText, "plan", "only the data-verdict part is the seam — text can never fake it");
const acceptOutOfOrder = phaseAt3([
user("teach me"),
verdict(true), // before any probe_done: out of order, advances nothing
]);
assert.equal(acceptOutOfOrder, "probe", "the chain never skips");
});
test("phaseAt3: a mere probe_done CALL (input-available, not yet concluded) keeps the probe open", () => {
const still = phaseAt3([user("teach me"), assistantTool("probe_done", "input-available")]);
assert.equal(still, "probe");
});
test("phaseAt3: legacy transcripts without probe_done read as probe (no backwards compat — data wiped)", () => {
const legacy = phaseAt3([user("teach me lacan"), assistantTool("plan_init", "output-available", PLAN_OK)]);
assert.equal(legacy, "probe", "plan_init alone never flips a phase post-split");
});
test("phaseTimeline: folds messages into phase spans for the dividers", () => {
const timeline = phaseTimeline([
user("teach me"),
assistantTool("quiz", "output-available"),
assistantTool("probe_done", "output-available", DONE_ASSESSMENT),
assistantTool("plan_init", "output-available", PLAN_OK),
user("what does node 1 mean?"), // side question — still planning
planAccepted(),
user("go on"),
]);
assert.deepEqual(timeline satisfies Phase3[], ["probe", "probe", "plan", "plan", "plan", "teach", "teach"]);
assert.equal(timeline.length, 7);
});
test("phaseAt3: empty transcript is probe", () => {
assert.equal(phaseAt3([]), "probe");
});