Wave A: pi-native tool layer — delete LhTool/toPiTool, TypeBox everywhere, terminate on questions

- src/agent/lh-tool.ts deleted: the zod-wrapped LhTool + toPiTool bridge
  (z.toJSONSchema per call, clientQuestion flag, hand-rolled result mapping)
  is gone. Tools are pi-agent-core native AgentTools now.
- New src/agent/tool-execute.ts owns what the wrapper added: the abort guard
  (runWithAbortGuard: run signal + 30s per-call timeout + 250ms grace,
  'Operation aborted' / timeout wording preserved), ToolInputError (the
  retryable '- problem at path' format), result mapping (string verbatim /
  object JSON / image parts, details rides the record's tool part output),
  and questionResult() — placeholder + terminate:true (owner Option A:
  belt of terminate + slim makeQuestionStopper with the clean-question
  rule; maxTurns deleted from the teaching turn entirely per owner).
- All 11 tool files + tools.ts registries converted to TypeBox parameters
  (typebox@1.3.7, pinned to pi-agent-core's copy). zod refinements
  (quizInputProblems, askOptionProblems, live-nodeId) moved from schemas
  into execute — the model sees the same retryable error shape.
- question-core + contracts/probe TypeBox-authored (single source shared
  with the UI); tool-parts parse seam runs the boundary's full chain
  (Convert → Check → Clean → refinement), fixing the live-smoke lockout:
  pi's boundary Value.Converts correctAnswer: 17 → '17' but the record
  keeps raw args, so a parse without Convert rejected a posed quiz
  (invisible pending question, turn ended). Pinned with the real smoke
  emission shape.
- Subagents converted: demo-builder/researcher AgentTool-native,
  visual-maker/pdf-reader declaration-only pi-ai Tools with local execute
  fns, extraction submit schemas re-authored in TypeBox (owner option 2:
  zod defaults → Type.Optional + fill at the submit seam).
- Accepted delta, flagged: TypeBox-native schemas skip pi's JSON-Schema
  primitive coercion fast path... (boundary still Value.Converts), and
  prober-quiz unknown keys ride through (Value.Clean strips at display).
- Tests: 988 passing (was 967 at the recon baseline); tool-abort and
  read-tool rewritten against the native seams; question-core/question-
  tools/tool-parts/demos/web-search/graph-tools/plan-tree/probe-done/
  tool-toggles converted to Value.Check/execute pins.

Live smoke (PORT=3004, real model): native ls turn streams end-to-end; a
real quiz pose (first attempt missing correctAnswer → tool-input-error →
model retry → posed) terminates the turn cleanly and persists pending.
This commit is contained in:
amosblomqvist
2026-09-25 23:32:13 +03:00
parent b07cb313f0
commit b87a6925af
35 changed files with 1357 additions and 1083 deletions
+42 -9
View File
@@ -10,6 +10,8 @@
import assert from "node:assert/strict";
import { describe, it } from "node:test";
import { quizInputProblems } from "../src/core/question-core.ts";
import { Value } from "typebox/value";
import {
parseAskInput,
parseProbeDoneInput,
@@ -59,15 +61,13 @@ describe("parseQuizInput", () => {
});
it("boundary parity: anything the TEACH schema accepts, the display parse accepts", () => {
const boundary = quizInputSchema.safeParse(TEACHER_QUIZ);
assert.ok(boundary.success, "fixture must be boundary-valid for this pin to mean anything");
assert.ok(parseQuizInput(boundary.data));
assert.ok(Value.Check(quizInputSchema, TEACHER_QUIZ), "fixture must be boundary-valid for this pin to mean anything");
assert.ok(parseQuizInput(TEACHER_QUIZ));
});
it("boundary parity: anything the PROBE schema accepts, the display parse accepts", () => {
const boundary = probeQuizInputSchema.safeParse(PROBER_QUIZ);
assert.ok(boundary.success);
assert.ok(parseQuizInput(boundary.data));
assert.ok(Value.Check(probeQuizInputSchema, PROBER_QUIZ));
assert.ok(parseQuizInput(PROBER_QUIZ));
});
it("the live crash shape — options as a JSON string — does not parse", () => {
@@ -75,6 +75,40 @@ describe("parseQuizInput", () => {
assert.equal(parseQuizInput(stringified), undefined);
});
it("boundary parity under Value.Convert: a NUMERIC correctAnswer parses (live-smoke lockout fix, 2026-09-26)", () => {
// The live smoke's model emitted correctAnswer: 17 (number). pi's
// boundary runs Value.Convert before Value.Check, so the quiz posed
// — but the record keeps the RAW args. A parse seam without Convert
// would reject exactly what the boundary accepted: the pending quiz
// renders nowhere while the turn has ended = learner lockout.
// Dock-visible parity case: the smoke's real emission — options whose
// values ARE the numbers, correctAnswer numeric. Convert → "17",
// refinement passes (it matches an option), the quiz parses.
const numeric = {
question: "What is 5 + 3 × 4?",
options: [
{ label: "17", value: "17" },
{ label: "27", value: "27" },
{ label: "9", value: "9" },
],
explanation: "Multiplication first.",
correctAnswer: 17,
};
const parsed = parseQuizInput(numeric);
assert.ok(parsed);
assert.equal(parsed.correctAnswer, "17", "converted to the string the boundary saw");
const numericArray = { ...numeric, multiSelect: true, correctAnswer: [17, "27"] };
const parsedArray = parseQuizInput(numericArray);
assert.ok(parsedArray);
assert.deepEqual(parsedArray.correctAnswer, ["17", "27"]);
// Non-convertible garbage still never parses (structure owns it)…
assert.equal(parseQuizInput({ ...PROBER_QUIZ, correctAnswer: { bad: true } }), undefined);
// …and Convert's widening (undefined → "") is caught by the refinement
// chain the boundary itself uses — garbage renders as a quiet fallback.
const widened = { ...PROBER_QUIZ, correctAnswer: undefined };
assert.equal(parseQuizInput(widened), undefined, "the refinement owns the reject");
});
it("garbage never parses: non-object, missing question, missing correctAnswer", () => {
assert.equal(parseQuizInput("a string"), undefined);
assert.equal(parseQuizInput({ options: [] }), undefined);
@@ -102,9 +136,8 @@ describe("parseProbeDoneInput", () => {
});
it("boundary parity: anything the tool schema accepts, the display parse accepts", () => {
const boundary = probeDoneInputSchema.safeParse(VALID_ASSESSMENT);
assert.ok(boundary.success);
assert.ok(parseProbeDoneInput(boundary.data));
assert.ok(Value.Check(probeDoneInputSchema, VALID_ASSESSMENT));
assert.ok(parseProbeDoneInput(VALID_ASSESSMENT));
});
it("the live crash shape — assessment STRINGIFIED — does not parse", () => {