Skip to content

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.

C++

A big message on a chat near the compaction threshold summarises first; the real-engine script can also run against a server that is already up

Niko12213ae6b3041b43
diff --git a/apps/host/scripts/context-e2e.mjs b/apps/host/scripts/context-e2e.mjs
index 2696f8880..02e87fe78 100644
--- a/apps/host/scripts/context-e2e.mjs
+++ b/apps/host/scripts/context-e2e.mjs
@@ -33,6 +33,8 @@ const MIN_FREE_GB = Number(env('E2E_MIN_FREE_GB', '7'));
const GGUF = join(DOCS, 'models', 'gyro-Gyro-S', 'Qwen3.8-Flash-Next-Gyro-S-TQ1_0.gguf');
const CUDA_BIN = env('E2E_CUDA_BIN', 'C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v13.0\\bin');
const WINDOW = 65536;
+/** A model server that is already running (the scripted worker, to try this script without a GPU): nothing is started or stopped. */
+const EXISTING = process.env.E2E_BASE_URL ?? '';
const sleep = ms => new Promise(r => setTimeout(r, ms));
const freeGb = () => {
@@ -43,8 +45,8 @@ const rows = [];
const note = (id, ok, detail) => { rows.push({ id, ok, detail }); console.log(`${ok ? 'PASS' : 'FAIL'} ${id} ${detail}`); };
if (!process.env.STRATA_HARNESS_DIR) { console.error('STRATA_HARNESS_DIR is not set'); process.exit(2); }
-for (const f of [ENGINE, GGUF]) if (!existsSync(f)) { console.error(`missing: ${f}`); process.exit(2); }
-if (freeGb() < 20) { console.error(`only ${freeGb()} GB of RAM are free; wait for the other streams (lock.py status) and run again`); process.exit(2); }
+if (!EXISTING) for (const f of [ENGINE, GGUF]) if (!existsSync(f)) { console.error(`missing: ${f}`); process.exit(2); }
+if (!EXISTING && freeGb() < 20) { console.error(`only ${freeGb()} GB of RAM are free; wait for the other streams (lock.py status) and run again`); process.exit(2); }
const scratch = mkdtempSync(join(tmpdir(), 'strata-ctx-e2e-'));
const configFile = join(scratch, 'server.json');
@@ -73,20 +75,19 @@ const finish = async (code) => {
};
process.on('SIGINT', () => finish(130));
-server = spawn(PYTHON, ['serve/server.py', '--engine', 'strata', '--config', configFile, '--port', String(PORT), '--no-api-key', '--ready-file', readyFile], {
+if (!EXISTING) server = spawn(PYTHON, ['serve/server.py', '--engine', 'strata', '--config', configFile, '--port', String(PORT), '--no-api-key', '--ready-file', readyFile], {
cwd: REPO, env: { ...process.env, PATH: `${CUDA_BIN};${process.env.PATH}`, STRATA_DATA_DIR: join(scratch, 'data') }, stdio: ['ignore', 'pipe', 'pipe'], windowsHide: true,
});
const sink = [];
-server.stdout.on('data', d => sink.push(String(d))); server.stderr.on('data', d => sink.push(String(d)));
-server.on('exit', code => { sink.push(`[server exited ${code}]`); server = null; });
+if (server) { server.stdout.on('data', d => sink.push(String(d))); server.stderr.on('data', d => sink.push(String(d))); server.on('exit', code => { sink.push(`[server exited ${code}]`); server = null; }); }
// Memory alarm: this run must never be what freezes the PC. Under the floor the engine is stopped (by PID) and the run ends.
-alarm = setInterval(() => { const free = freeGb(); if (free < MIN_FREE_GB) { console.error(`ALARM: ${free} GB of RAM free, stopping the engine`); finish(3); } }, 3000);
-alarm.unref?.();
+if (!EXISTING) alarm = setInterval(() => { const free = freeGb(); if (free < MIN_FREE_GB) { console.error(`ALARM: ${free} GB of RAM free, stopping the engine`); finish(3); } }, 3000);
+alarm?.unref?.();
-const base = `http://127.0.0.1:${PORT}`;
+const base = EXISTING || `http://127.0.0.1:${PORT}`;
let ready = false;
for (let i = 0; i < 120 && !ready; i++) {
- if (!server) break;
+ if (!EXISTING && !server) break;
try { ready = (await fetch(`${base}/health`, { signal: AbortSignal.timeout(2000) })).ok; } catch { /* still loading */ }
if (!ready) await sleep(2000);
}
@@ -137,7 +138,7 @@ const needle = 'NOTE FROM OPS: the vault code is COBALT-27. Do not lose it.';
const timeline = [];
let notes = 0, refusedAny = false, peak = 0;
{
- for (let k = 0; k < 6; k++) {
+ for (let k = 0; k < 5; k++) {
const body = logText(34_000, k * 1000, k === 0 ? needle : '');
const t = await turnsOf(`${body}\n\nReply with the single word OK.`);
const note = noteOf(t.frames);
@@ -149,7 +150,7 @@ let notes = 0, refusedAny = false, peak = 0;
console.log(' ' + timeline.at(-1));
if (t.last.type === 'error') break;
}
- note('B six dense messages (about 14K tokens each) in a 64K window: no error', !refusedAny, `peak context ${peak} of ${WINDOW}; ${notes} compaction note(s) in the chat`);
+ note('B five dense messages (about 14K tokens each) in a 64K window: no error', !refusedAny, `peak context ${peak} of ${WINDOW}; ${notes} compaction note(s) in the chat`);
note('B compaction ran and was announced', notes >= 1, timeline.join(' | '));
}
@@ -175,7 +176,7 @@ let notes = 0, refusedAny = false, peak = 0;
rows.push({ id: 'D the needle', ok: /COBALT-27/.test(answer), detail: `reported only: answer "${answer.slice(0, 80)}"` });
}
-console.log('\n' + rows.map(r => `${r.ok ? 'PASS' : 'FAIL'} ${r.id}`).join('\n'));
+console.log('\n' + rows.map(r => `${r.ok ? 'PASS' : r.id.startsWith('D') ? 'INFO' : 'FAIL'} ${r.id}${r.id.startsWith('D') ? ` ${r.detail}` : ''}`).join('\n'));
const failed = rows.filter(r => !r.ok && !r.id.startsWith('D')).length;
writeFileSync(join(scratch, 'result.json'), JSON.stringify(rows, null, 2));
await finish(failed ? 1 : 0);
diff --git a/apps/host/server/agent/agent-bridge.mjs b/apps/host/server/agent/agent-bridge.mjs
index 7f61de553..4a0ccbc83 100644
--- a/apps/host/server/agent/agent-bridge.mjs
+++ b/apps/host/server/agent/agent-bridge.mjs
@@ -22,7 +22,7 @@ import { randomUUID } from 'node:crypto';
import { GatewayClient } from './gateway-client.mjs';
import { HarnessProcess, resolveHarness, redact, stopStaleAgent } from './harness-runtime.mjs';
import { StatusReporter } from './status.mjs';
-import { buildProfile, routeFor, MODE_PRESETS, modeOf, effortOf } from './profile.mjs';
+import { buildProfile, routeFor, MODE_PRESETS, modeOf, effortOf, localOutputBudget, compactionHeadroom } from './profile.mjs';
import { TurnProjector, parseArguments as parseArgs } from './events.mjs';
import { ChildProjector } from './children.mjs';
import { nearestBudget } from './llm-shape.mjs';
@@ -1203,11 +1203,14 @@ export function createAgentBridge({ dataDir, componentsDir = null, bundledDir =
const now = fresh ? null : await contextOf(runtime, sessionId);
const window = now?.window || route.contextWindow;
const pressure = now?.tokens ?? 0;
- let verdict = decideFit({ window, pressure, messageTokens: tokens });
+ // Where the Harness starts summarising by itself on this window (profile.mjs: the reply keeps its room, then the headroom).
+ const reply = localOutputBudget(window);
+ const threshold = Math.floor(Math.min(window * 0.8, window - reply - compactionHeadroom(window, reply)));
+ let verdict = decideFit({ window, pressure, messageTokens: tokens, threshold });
// The earlier conversation carried into a fresh session is what pushed it over: send the message alone instead.
if (verdict.action === 'too-large' && carried !== message) {
const alone = await count(message);
- if (alone != null) { tokens = alone; verdict = decideFit({ window, pressure, messageTokens: alone }); }
+ if (alone != null) { tokens = alone; verdict = decideFit({ window, pressure, messageTokens: alone, threshold }); }
}
return { ...verdict, tokens, window, pressure };
}
diff --git a/apps/host/server/agent/context-fit.mjs b/apps/host/server/agent/context-fit.mjs
index 04552cb29..0abcb08a2 100644
--- a/apps/host/server/agent/context-fit.mjs
+++ b/apps/host/server/agent/context-fit.mjs
@@ -17,17 +17,23 @@ export const COUNT_ABOVE_CHARS = 3000;
export const replyRoom = window => Math.max(512, Math.min(REPLY_FLOOR, Math.floor(window / 8)));
/**
- * @param {{window:number, pressure?:number, base?:number, messageTokens:number}} input
- * `pressure` is what the chat holds now (measured), `base` what an empty chat holds, `messageTokens` the new message.
+ * @param {{window:number, pressure?:number, base?:number, messageTokens:number, threshold?:number}} input
+ * `pressure` is what the chat holds now (measured), `base` what an empty chat holds, `messageTokens` the new message, `threshold` where the
+ * Harness starts summarising by itself. The Harness looks at that only before its NEXT step, so a big message that arrives now would otherwise
+ * be sent on top of a chat already near the threshold (2026-10-22: 54K of 64K held by the time it was looked at): summarise first.
* @returns {{action:'send'|'summarize'|'too-large', room:number, needs:number}}
*/
-export function decideFit({ window, pressure = 0, base: given = BASE_TOKENS, messageTokens }) {
+export function decideFit({ window, pressure = 0, base: given = BASE_TOKENS, messageTokens, threshold = 0 }) {
const reply = replyRoom(window);
const base = Math.min(given, Math.floor(window * 0.6)); // on a small window the first prompt cannot be what the window is made of
// A chat holds at least what an empty one does (the system prompt and the tools), so a new chat is not counted as empty.
const needs = Math.max(pressure, base) + messageTokens + reply;
if (!(window > 0) || !(messageTokens >= 0)) return { action: 'send', room: Infinity, needs: 0 };
- if (needs <= window) return { action: 'send', room: window - needs, needs };
+ if (needs <= window) {
+ const worthIt = pressure > base + window * 0.1; // older messages worth summarising: more than a tenth of the window beyond the first prompt
+ if (threshold > 0 && worthIt && pressure + messageTokens > threshold) return { action: 'summarize', room: window - needs, needs, why: 'threshold' };
+ return { action: 'send', room: window - needs, needs };
+ }
// Not as it stands. Does the message fit in an otherwise empty chat (only the system prompt and the tools before it)?
if (base + messageTokens + reply <= window) return { action: 'summarize', room: window - (base + messageTokens + reply), needs };
return { action: 'too-large', room: window - (base + messageTokens + reply), needs };
diff --git a/apps/host/tests/agent-context.test.mjs b/apps/host/tests/agent-context.test.mjs
index 923f695d4..600b1272f 100644
--- a/apps/host/tests/agent-context.test.mjs
+++ b/apps/host/tests/agent-context.test.mjs
@@ -162,6 +162,17 @@ test('fit: a message that fits is sent, one that fits after a summary is summari
assert.equal(decideFit({ window: 8192, pressure: 0, messageTokens: 5000 }).action, 'too-large', 'a small window has less room for a message');
});
+test('fit: a big message on a chat already near the compaction threshold summarises first, so the window is never come close to', () => {
+ // 64K window: the Harness starts summarising at 40,142 (profile.mjs), but looks only before its NEXT step; a message that arrives now is not in that count.
+ const at = { window: 65536, threshold: 40142, messageTokens: 14400 };
+ const near = decideFit({ ...at, pressure: 39840 });
+ assert.equal(near.action, 'summarize'); assert.equal(near.why, 'threshold');
+ assert.equal(decideFit({ ...at, pressure: 15000 }).action, 'send', 'still under the threshold with the message on top');
+ assert.equal(decideFit({ ...at, pressure: 13000, messageTokens: 30000 }).action, 'send', 'a chat that holds little beyond its first prompt has nothing worth summarising');
+ assert.equal(decideFit({ window: 65536, messageTokens: 14400, pressure: 39840 }).action, 'send', 'no threshold given: only the window decides');
+ assert.equal(decideFit({ ...at, pressure: 59048, messageTokens: 5000 }).why, undefined, 'a refusal-sized case is the old rule, not the threshold');
+});
+
test('count: the engine\'s own number, or nothing when it cannot be had (then nothing is blocked)', async () => {
const sent = [];
const ok = async (url, init) => { sent.push([url, JSON.parse(init.body), init.headers]); return { ok: true, json: async () => ({ input_tokens: 70141 }) }; };