Files
pepa-pi-bot/runtime/coach/postmortem.test.js
T
mayatnikovandClaude Opus 4.7 0d97ccccfa feat(v0.3.0): paradigm shift — TimeWeb-only LLM + persistent advisor trail + improvement queue
This is the rc.4 batch the user requested:

  1. Emergency triggers (low HP + close hostile, lava-under-foot)
     bypass the long cooldown so the LLM is consulted BEFORE the bot
     dies, not after.
  2. Active manifesto need is now included in the advisor user prompt
     — the LLM picks suggestions that satisfy the bot's current
     concrete need (L2 tools_wood → "gather logs nearby" not
     "explore further").
  3. Every advisor recommendation is persisted to SQLite
     (advisor_recommendations table) with full token usage. The
     reflex marks 'applied=1' when it dispatches and updates
     outcome_ok/code when the dispatch completes. Ground truth for
     "is the LLM actually helping" lives in the DB, not in logs.
  4. Pi CLI is OUT of every background loop. coach/postmortem and
     coach/reflect now go through the same TimeWeb endpoint
     fast-advisor uses, via the shared coach/llm-call.js helper.
     Pi is reserved for manual operator commands.
  5. The LLM (postmortem, reflect, advisor) can flag "structural
     gaps" — missing skills/features the operator should implement.
     These land in the new improvement_requests table. Dedup by
     title bumps `votes` instead of inserting duplicates so the
     queue doesn't bloat. Operator views via
     `node scripts/list-improvements.js`.
  6. A deterministic trigger-tuner runs hourly: reads 24h of
     recommendation stats, flags triggers whose success rate is
     below 25% (sample ≥ 5) or whose prompts are expensive (>1000
     input tokens) with mediocre payoff. Improvements get
     source="tuner", category="tuning". No LLM call.

New files:
  runtime/coach/llm-call.js        — askAnalytical() helper
  runtime/coach/trigger-tuner.js   — stats → improvements
  runtime/coach/trigger-tuner.test.js
  scripts/list-improvements.js     — operator CLI

Schema additions:
  advisor_recommendations: id, ts, trigger_reason, planned_skill,
    recommended_skill, action, rationale, active_need, tokens_in,
    tokens_out, latency_ms, applied, outcome_ok, outcome_code, outcome_at
  improvement_requests: id, ts, source, category, title, description,
    context, priority, status, duplicate_of, votes, implemented_at, notes

Renamed env-var consumers:
  Pi-coach drainOnce({ askPi })   → drainOnce({ askAnalyticalFn? })
  Pi-reflect runOnce({ askPi })   → runOnce({ askAnalyticalFn? })
  bot.js attachCoach/attachReflect no longer pass askPi
  attachTuner() added to bot.js spawn handler
  lessons.source 'pi-coach'   → 'timeweb-coach'
  lessons.source 'pi-reflect' → 'timeweb-reflect'

Token cost measured live:
  ~705 input + 45 output = ~750 total per advisor call
  worst case @ 6 calls/hour rate cap = ~108K tokens/day
  OpenAI gpt-5-mini reference price: ~$0.60/month

Operator usage:
  node scripts/list-improvements.js                # open queue
  node scripts/list-improvements.js --stats        # advisor performance
  node scripts/list-improvements.js --done 17 "shipped in 0.3.1"
  node scripts/list-improvements.js --reject 18 "duplicate"

Tests: 360 green (was 332, +28 new).

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-05-27 19:12:48 +03:00

172 lines
6.6 KiB
JavaScript

import { test } from "node:test";
import assert from "node:assert/strict";
import { mkdtempSync, rmSync, writeFileSync } from "node:fs";
import { tmpdir } from "node:os";
import { join } from "node:path";
import { initKnowledge, unanalysedDeaths, recall } from "../knowledge/index.js";
import { closeStore, __resetForTests, isAvailable } from "../knowledge/store.js";
import { attach, detach, drainOnce, __testing } from "./postmortem.js";
const { captureDeath, extractJson, inferCause, buildPrompt } = __testing;
function mockBot({ pos = { x: 100, y: 64, z: 200 }, hp = 0, food = 14, entities = {} } = {}) {
const handlers = {};
return {
entity: { position: { ...pos, distanceTo(o) { return Math.hypot(pos.x - o.x, pos.z - o.z); } } },
health: hp,
food,
time: { timeOfDay: 18000 },
entities,
isRaining: false,
inventory: { items: () => [] },
on(ev, fn) { handlers[ev] = fn; },
emit(ev, payload) { handlers[ev]?.(payload); },
};
}
test("inferCause: hostile present → 'hostile'", () => {
assert.equal(inferCause({ bot: { food: 20, entity: { position: { y: 64 } } }, hostile: "creeper" }), "hostile");
assert.equal(inferCause({ bot: { food: 0, entity: { position: { y: 64 } } }, hostile: null }), "starvation");
assert.equal(inferCause({ bot: { food: 20, entity: { position: { y: 20 } } }, hostile: null }), "fall");
assert.equal(inferCause({ bot: { food: 20, entity: { position: { y: 64 } } }, hostile: null }), "unknown");
});
test("extractJson: tolerates fences and surrounding text", () => {
assert.deepEqual(extractJson('```json\n{"a": 1}\n```'), { a: 1 });
assert.deepEqual(extractJson('Reply: {"a": 2} done'), { a: 2 });
assert.equal(extractJson("not json"), null);
assert.equal(extractJson(""), null);
});
test("buildPrompt: returns {system, user}, includes all death rows + improvements schema", () => {
const rows = [
{ id: 1, ts: Date.now(), x: 100, y: 64, z: 200, cause: "hostile", hostile: "creeper", last_skill: "gather.logs", last_skill_code: "timeout", food_at_death: 14, context_blob: JSON.stringify({ recentScenarios: [{ skillId: "gather.logs", code: "timeout" }] }) },
{ id: 2, ts: Date.now(), x: 102, y: 64, z: 201, cause: "hostile", hostile: "creeper", last_skill: "explore.far", last_skill_code: "done", food_at_death: 12, context_blob: null },
];
const { system, user } = buildPrompt(rows);
assert.match(user, /death id=1/);
assert.match(user, /death id=2/);
assert.match(user, /creeper/);
assert.match(system, /Reply with ONE JSON object/);
assert.match(system, /improvements/);
assert.match(system, /Valid skill ids/);
});
test("captureDeath: builds a row with context blob and inferred cause", () => {
const bot = mockBot({ entities: {
1: { type: "hostile", name: "zombie", position: { x: 101, y: 64, z: 200, distanceTo: (p) => Math.hypot(101 - p.x, 200 - p.z) } },
}});
const stateDir = mkdtempSync(join(tmpdir(), "pepa-coach-test-"));
writeFileSync(join(stateDir, "current-task.json"), JSON.stringify({ label: "gather.logs", lastCode: "timeout" }));
const death = captureDeath(bot, { stateDir });
assert.equal(death.cause, "hostile");
assert.equal(death.hostile, "zombie");
assert.equal(death.lastSkill, "gather.logs");
assert.equal(death.lastSkillCode, "timeout");
assert.equal(death.x, 100);
assert.ok(death.contextBlob.snapshot);
rmSync(stateDir, { recursive: true, force: true });
});
test("attach + emit('death'): inserts row in knowledge DB", async () => {
const stateDir = mkdtempSync(join(tmpdir(), "pepa-coach-test-"));
__resetForTests();
await initKnowledge({ stateDir });
if (!isAvailable()) {
// Without sqlite the no-op contract is enough.
assert.ok(true);
rmSync(stateDir, { recursive: true, force: true });
return;
}
const bot = mockBot();
attach(bot, { stateDir });
bot.emit("death");
// Insert is sync.
const pending = unanalysedDeaths({ limit: 10 });
assert.ok(pending.length >= 1, "death row inserted");
detach();
closeStore();
rmSync(stateDir, { recursive: true, force: true });
});
test("drainOnce: respects budget and parses analytical LLM reply (incl. improvements)", async () => {
const stateDir = mkdtempSync(join(tmpdir(), "pepa-coach-test-"));
__resetForTests();
await initKnowledge({ stateDir });
if (!isAvailable()) {
assert.ok(true);
rmSync(stateDir, { recursive: true, force: true });
return;
}
// Seed one unanalysed death.
const bot = mockBot();
attach(bot, { stateDir });
bot.emit("death");
detach();
const lessonsBefore = recall({ category: "combat" }).length;
// TimeWeb path needs env vars to satisfy the llmAvailable check.
const prevKey = process.env.TIMEWEB_API_KEY;
const prevModel = process.env.TIMEWEB_MODEL;
process.env.TIMEWEB_API_KEY = "test-key";
process.env.TIMEWEB_MODEL = "test-model";
const fakeReply = {
cause: "creeper_explosion_unarmed",
next_action: "shelter at dusk",
lessons: [{
lesson: "Stop attacking creepers without armour; dig down 2 instead.",
category: "combat",
trigger_hostile: "creeper",
avoid_skill: "attack creeper",
prefer_skill: "survive.flee",
confidence: 0.85,
}],
improvements: [
{ title: "Add craft.shield skill", description: "No skill to craft a shield when creepers are around.", category: "skill", priority: 2 },
],
};
const askAnalyticalFn = async () => fakeReply;
const result = await drainOnce({ stateDir, force: true, askAnalyticalFn });
if (prevKey === undefined) delete process.env.TIMEWEB_API_KEY;
else process.env.TIMEWEB_API_KEY = prevKey;
if (prevModel === undefined) delete process.env.TIMEWEB_MODEL;
else process.env.TIMEWEB_MODEL = prevModel;
assert.equal(result.ok, true);
assert.equal(result.analysed, 1);
assert.equal(result.lessons, 1);
assert.equal(result.improvements, 1);
const after = recall({ hostile: "creeper", category: "combat" });
assert.ok(after.length > lessonsBefore, "new lesson recorded");
const stillPending = unanalysedDeaths({ limit: 10 });
assert.equal(stillPending.length, 0, "death marked analysed");
closeStore();
rmSync(stateDir, { recursive: true, force: true });
});
test("drainOnce: skipped when LLM not configured", async () => {
const stateDir = mkdtempSync(join(tmpdir(), "pepa-coach-test-"));
__resetForTests();
await initKnowledge({ stateDir });
if (!isAvailable()) {
rmSync(stateDir, { recursive: true, force: true });
return;
}
const prevKey = process.env.TIMEWEB_API_KEY;
delete process.env.TIMEWEB_API_KEY;
const result = await drainOnce({ stateDir, force: true });
if (prevKey !== undefined) process.env.TIMEWEB_API_KEY = prevKey;
assert.equal(result.ok, false);
assert.equal(result.reason, "llm not configured");
closeStore();
rmSync(stateDir, { recursive: true, force: true });
});