Почему Hermes Agent нужен режим 7×24: трёхуровневая память и замеры ресурсов на Mac Mini M4 (2026)

3 слоя памяти · reboot ≠ wipe · матрица Pi/VPS/M4 · TCO 24 мес · 6-шаговый runbook

Hermes Agent трёхуровневая память и Mac Mini M4 7×24

Hermes Agent «умнеет» не магией, а трёхуровневой памятью на диске: USER.md ~1 375 символов профиля, MEMORY.md cap ~2 200 символов на Skill, SQLite FTS5 для full-text retrieval — Skill пишется только после 5+ tool calls в завершённой задаче. Разбор архитектуры, почему нужен 7×24 uptime, бенчмарк Pi / VPS / Mac Mini M4 с измеренными RAM band, TCO 24 месяца и 6-шаговый runbook. Reboot не стирает persisted memory; sleeping Gateway рвёт каналы и Skill polish loop.

01

Трёхуровневая память: USER.md, MEMORY.md и SQLite FTS5

Многие видят Nous Research Hermes Agent как chat shell с tools. Persistence layer глубже. Tier 1: session context — tool state и reasoning trace в RAM, gone on restart. Tier 2: Skill Documents, markdown playbooks после complex tasks, dedup в data directory — community cap ~2 200 символов. Tier 3: persistent user model в USER.md, budget ~1 375 символов.

Retrieval через SQLite FTS5 по Skills и memory entries. Agent query local index перед context injection — дешевле по tokens, значит disk IO и index health так же критичны, как compute. Skill synthesis gate: минимум пять tool calls на task; хост не должен suspend mid-chain.

Компонент памятиРазмер / механизмSurvives reboot?Зачем 7×24
Session contextCurrent turn + tool stateНет — reconnectGateway live; IM webhook timeout = broken chain
USER.md~1 375 символов профильДа — on diskMigration = data dir copy; sleep тормозит profile iteration
MEMORY.md / Skills~2 200 символов на entryДа — on diskFTS5 растёт с writes; backup non-optional
SQLite FTS5Local full-text indexДа — DB fileDisk jitter / VPS IO cap = retrieval latency

Reboot ≠ memory wipe только для flushed layers. Channel UX, cron и in-flight 5+ tool chains всё равно ломаются. Субъективный 30-day narrative: 30 дней с Hermes; здесь — architecture и resource math.

  1. 01

    Думать, reboot стирает всё: Skills и USER.md в data directory — без backup host swap = amnesia feeling.

  2. 02

    Игнорировать 5+ tool gate: short chats не становятся Skills; sleep mid-task = extraction loop never completes.

  3. 03

    FTS5 как black box: corrupt index / full disk → «писал, но не find» — мониторить data dir size и SQLite health.

  4. 04

    Не trim USER.md: 1 375-char budget конечен; bloated profile dilutes preference weights.

  5. 05

    Split Gateway от model host: dead Gateway + live cloud backend = IM drop — 7×24 = whole chain up.

02

Зачем Hermes Agent 7×24: Gateway, каналы и Skill compounding

Hermes = always-on agent: 20+ каналов (Telegram, Discord, Slack…) через webhooks в Gateway; cron будит subtasks; Honcho медленно refresh user model. Любой link down → missed callbacks, queue backlog, delayed Skill writes — субъективно new assistant каждую неделю, хотя Skill files растут на disk.

7×24 не ops theater — match time axis трёх tiers. Session layer хочет ms response; Skill layer — long tasks с 5+ tools; user model — weeks compounding. Closed laptop, intermittent NAS, VPS throttled neighbor IO — compounding curve flatten. Dedicated host = process survival, stable net, predictable disks как SLA, не gamble на power plug.

Memory compounds on disk, но feeling smarter = Gateway без пропусков смен — gap между 7×24 и «запущу когда вспомню».

Minimum acceptance bar для always-on node

  • Process: Gateway + backend 30 days без manual restart (OS updates — change window).
  • Channels: любой IM, dispatch task, no timeout 24h; cron on schedule.
  • Data: writable data dir, healthy FTS5; weekly backup restores USER.md + Skills.
  • Resources: ≥20% RAM headroom — OOM не kill Gateway mid-chain.
03

Raspberry Pi, VPS, Mac Mini M4: benchmark matrix Hermes

Тот же curl -fsSL https://get.hermes-agent.org | bash — разница в memory bandwidth, disk IO, macOS-native path. Таблица: qualitative band для Gateway + Telegram + local Ollama Hermes-3 8B (intermittent inference). Exact numbers shift с quantization и channel count — для review meeting, не lab cert.

HostIdle RAMPeak RAMCPU / powerHermes fit
Raspberry Pi 5 · 8GB≈1.5 GB headroomGateway alone ≈4 GB; local 8B not viableLow-power ARM; SD IO bottleneckAPI-only gateway; weak Skill compounding
Linux VPS 4C8G≈5 GB usableAPI ≈6 GB; Docker +2 GBShared vCPU; IOPS capRemote SSH OK; no macOS — awkward Skills
Mac Mini M4 16GB≈9 GB usableLocal 8B + channels ≈14–15 GB at ceilingIdle ≈12W; burst 25–35WNative macOS; single channel + local at limit
Mac Mini M4 32GB≈22 GB usable8B + dual channel + cron ≈18–20 GBSame silicon, less pressureProduction pick — room for Skill + FTS5 growth

Unified Memory (UMA) на M4 режет CPU↔GPU copies при local inference; macOS — shortest installer + Ollama path. Pi saves watts, не держит 8B; VPS cheap rent, теряет time на RTT и IO — когда Skills + FTS5 в GB, важнее stable disk latency, чем экономия на month one.

04

TCO 24 месяца: купить Mac Mini M4 vs monthly rental

Decision object: total cost run memory compounding 24 months — hardware, power, ops hours, upgrade anxiety, migration. Rental = CapEx → OpEx; для prod Skills + channels часто lower decision cost vs buy-plus-self-support.

TCO dimension (24 мес)Buy M4 16GBRent M4 32GB
Hardware cash flowUpfront + tax; self-model depreciationFixed monthly × 24; RAM upgrade без box swap
Power (7×24)≈12–35W × 24h × 730 daysIn service fee
Ops hoursWarranty, OS, outage on youHardware swap; remote KVM
Hermes data assetsUSER.md / Skills / FTS5 bound to one machineBackup → restore; wipe on return
Upgrade riskM-series cadence tempts second buyNew spec at contract end
Opportunity costHardware research vs Skill polishFocus agent workflows + channels

6-шаговый runbook: lease → FTS5 smoke test

  1. 01

    Pick RAM: API-only + one channel — 16GB; local Hermes-3 + multi-channel + cron — 32GB.

  2. 02

    Order & access: lease ID, remote path; MDM для org.

  3. 03

    Acceptance: Apple Silicon, disk ≥256GB, macOS on official Hermes path; disable sleep.

  4. 04

    Install Hermes: official one-liner, hermes init; data dir + backup policy.

  5. 05

    7×24 smoke: bind IM, long task 5+ tools; after 24h check Skill write + FTS5 retrieval.

  6. 06

    Backup & off-board: export data dir; before lease end migrate USER.md / Skills + wipe per policy.

bash · macOS
curl -fsSL https://get.hermes-agent.org | bash
hermes init
hermes model

Tip: Pin Hermes version в prod; после hermes model switch — watch 24h memory curve перед second IM channel.

05

Citable params и next steps

  • USER.md budget: ~1 375 символов — trim when over budget.
  • MEMORY.md / Skill cap: ~2 200 символов per entry + dedup.
  • Skill trigger: ≥5 tool calls per task.
  • Search: local SQLite FTS5 — plan disk + backup as data dir grows.
  • Measured RAM: Gateway idle 200–400MB; Ollama Hermes-3 8B peaks 8–12GB; prod comfort 32GB UMA.
  • Power band: M4 7×24 idle ~12W, inference burst 25–35W.

Moat Hermes = three-tier memory compounding on disk — но curve нужен 7×24 Gateway, healthy FTS5, enough UMA finish 5+ tool chains. Pi/VPS pass installer, trim Skill curve on local inference или IO; Mac Mini M4 rental = hardware as predictable service.

VpsMesh Mac Mini M4 monthly rental: 16/32GB UMA, remote access, wipe-on-return. Цены: цены аренды Mac Mini M4, deploy/FAQ: центр помощи, order: оформить заказ.

Warning: Не migrate hosts + rebuild FTS5 + wipe Skill dir в один weekend — root-cause impossible. Сначала 24h Gateway stable, потом model routing или bulk memory import.

FAQ

Топ-3 вопроса

Нет. Skill Documents, USER.md, MEMORY.md, SQLite FTS5 on disk survive reboot; только session context breaks. Нужен 7×24 stable host + backups — sleeping laptop drops channels и long tool chains.

Gateway idle ~200–400MB; local Ollama Hermes-3 8B peaks 8–12GB. Parallel channels + cron — 16GB tight, 32GB Unified Memory safer. Tiers: страница цен.

Если Skill compound + channel uptime > owning silicon, 24-month rental = fixed OpEx — often beats buy+ops. Order: оформить заказ; setup: центр помощи.